JP7259110B2 - Information processing device, information processing method, and information processing program - Google Patents

Description

The present invention relates to an information processing device, an information processing method, and an information processing program.

Conventionally, techniques for generating information indicating whether or not a user performs a predetermined action have been provided. For example, a technology is provided that visually displays a portion of content that is likely to be clicked by the user.

JP 2014-174805 A

However, with the above-described conventional technology, it is not always possible to appropriately generate information indicating the relationship between the context regarding content distribution and the content. For example, it is difficult to appropriately generate information indicating the relationship between the content distribution context and the content only by visibly displaying portions that are likely to be clicked by the user.

The present application has been made in view of the above, and an object thereof is to provide an information processing device, an information processing method, and an information processing program that appropriately generate information indicating the relationship between content and context regarding distribution of content. and

An information processing apparatus according to the present application includes an acquisition unit that acquires context information related to distribution of content and content information related to the content, and based on the context information and the content information acquired by the acquisition unit, the content relationship information indicating a combination of the first element of the context information and the second element of the content information related to the distribution of the content information, wherein the target user corresponding to the first element is combined with the first element and a generation unit that generates relationship information used to provide target content corresponding to the second elements, and the acquisition unit generates the plurality of first elements of the context information and the plurality of content information. Acquiring a plurality of combination information indicating each combination of the second elements, and generating the relationship information indicating the combination of the first element and the second element based on the plurality of combination information It is characterized by

According to one aspect of the embodiment, it is possible to appropriately generate information indicating the relationship between the context regarding the distribution of content and the content.

FIG. 1 is a diagram illustrating an example of information processing according to an embodiment. FIG. 2 is a diagram illustrating a configuration example of an information processing system according to the embodiment; FIG. 3 is a diagram illustrating a configuration example of an information processing apparatus according to the embodiment; 4 is a diagram illustrating an example of a user information storage unit according to the embodiment; FIG. 5 is a diagram illustrating an example of a behavior information storage unit according to the embodiment; FIG. 6 is a diagram illustrating an example of a distribution information storage unit according to the embodiment; FIG. 7 is a diagram illustrating an example of a model information storage unit according to the embodiment; FIG. 8 is a diagram illustrating an example of a prediction information storage unit according to the embodiment; FIG. 9 is a diagram illustrating an example of a relationship information storage unit according to the embodiment; FIG. FIG. 10 is a flowchart illustrating an example of information processing according to the embodiment; FIG. 11 is a diagram illustrating an example of information processing according to the embodiment; FIG. 12 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing apparatus.

Hereinafter, modes for implementing an information processing apparatus, an information processing method, and an information processing program (hereinafter referred to as "embodiments") according to the present application will be described in detail with reference to the drawings. The information processing apparatus, information processing method, and information processing program according to the present application are not limited to this embodiment. Also, in each of the following embodiments, the same parts are denoted by the same reference numerals, and overlapping descriptions are omitted.

(embodiment)
[1. information processing]
An example of information processing according to the embodiment will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of information processing according to an embodiment. In the example of FIG. 1, the information processing apparatus 100 (see FIG. 3) uses a model generated based on the behavior information of the first user located in a predetermined area AR11 (area A) to generate another area AR21 (area B). ) to generate information about a second user located at . In the example of FIG. 1, push notification is shown as one form of content distribution, but content distribution may be in various forms. For example, content distribution in the information processing system 1 is not limited to push notification, and may be so-called pull-type content distribution in which content is requested from the terminal device 10 . For example, the push notification may be information output from the terminal device 10 regardless of the operation of the terminal device 10 by the user. For example, push notification may be to actively notify the user of predetermined information. Further, hereinafter, an operation of displaying information on the push notification on the terminal device 10 by selecting or specifying the push notification may be referred to as an opening operation. Also, the size and shape of the area referred to here may be appropriately set according to the purpose of processing. For example, areas of various sizes such as "prefecture", "city", "ward", and "town" may be appropriately set.

[Configuration of information processing system]
Prior to the description of FIG. 1, the configuration of the information processing system 1 will be described with reference to FIG. FIG. 2 is a diagram illustrating a configuration example of an information processing system according to the embodiment; As shown in FIG. 2 , the information processing system 1 includes a terminal device 10 , an information providing device 50 and an information processing device 100 . The terminal device 10, the information providing device 50, and the information processing device 100 are connected via a predetermined network N so as to be communicable by wire or wirelessly. The information processing system 1 shown in FIG. 2 may include multiple terminal devices 10 , multiple information providing devices 50 , and multiple information processing devices 100 .

The terminal device 10 is an information processing device used by a user. The terminal device 10 accepts various operations by the user. In addition, below, the terminal device 10 may be described as a user. That is, hereinafter, the user can also be read as the terminal device 10 . Note that the terminal device 10 described above is implemented by, for example, a smartphone, a tablet terminal, a notebook PC (Personal Computer), a desktop PC, a mobile phone, a PDA (Personal Digital Assistant), or the like. FIG. 1 shows a case where the terminal device 10 is a smart phone.

Also, the terminal device 10 has a function such as a GPS (Global Positioning System) sensor, etc., and is capable of detecting and acquiring a position. For example, the terminal device 10 transmits action information such as position information to the information processing device 100 . In addition, the terminal device 10 may have various sensors without being limited to the above. For example, the terminal device 10 acquires biological information such as an audio sensor (microphone), acceleration sensor, gyro sensor, temperature sensor, humidity sensor, illuminance sensor, pressure sensor, proximity sensor, odor, sweat, heartbeat, pulse, and brain wave. It may have various sensors, such as a sensor for Then, the terminal device 10 may transmit various sensor information detected by various sensors to the information processing device 100 . For example, the terminal device 10 may detect user's speech information (sensor information) with a voice sensor and transmit the detected speech information (sensor information) to the information processing device 100 .

For example, the terminal device 10 may have various functions such as a temperature sensor and an air pressure sensor, and may be capable of detecting and acquiring information around the user, such as temperature and air pressure. In addition, the terminal device 10 may have various functions such as a heart rate sensor, and may be capable of detecting and acquiring biological information of the user. For example, a user using the terminal device 10 may be able to acquire his/her own biometric information using the terminal device 10 by wearing a wearable device capable of communicating with the terminal device 10 . For example, a user using the terminal device 10 wears a wristband-type wearable device that can communicate with the terminal device 10, so that the terminal device 10 acquires information about the user's own heartbeat (pulse). It may be possible.

The context information related to content distribution (hereinafter also simply referred to as “context information”) includes information indicating various contexts related to content distribution. For example, the context information includes information indicating the context of the user to whom the content is delivered and information indicating the context of the user's location. For example, context information may include information about user attributes. For example, context information may include information indicating what kind of user the user is. For example, contextual information may include information about various users such as age, profile, interests, and so on. For example, contextual information may include information about the context of time and location. For example, the context information may include information indicating what kind of situation the time and place were when the push notification was performed. For example, context information may include information about time of day. For example, the context information may include information indicating which of the 24 time periods from 0:00 to 23:00 corresponds to the time period. For example, context information may include four time zones: morning (4:00 to 10:00), afternoon (10:00 to 16:00), evening (16:00 to 22:00), and midnight (22:00 to 4:00). may include information indicating which one of For example, the context information may include information on POIs (Points of Interest) indicating facility names such as stores and stadiums corresponding to the user's location. For example, the context information may include information related to various times and places such as how long the user stayed at the place, location attributes (commercial area, residential area, etc.), and time attributes (morning, night, etc.). Information may be included.

For example, contextual information may include information about the context of the environment. For example, the context information may include information indicating what kind of environment the user is in when the push notification is made. For example, the context information may include various information (surrounding information) such as weather, user's emotion (mood), temperature, atmospheric pressure, and the like. For example, contextual information may include information indicating whether the weather corresponds to sunny, cloudy, rainy, or snowy. For example, context information may include information indicating whether the user's mood is positive or negative. For example, the context information may include information indicating whether the user's mood is a positive emotion such as joy or fun, or a negative emotion such as anger or sadness.

For example, the information processing apparatus 100 estimates the user's emotion (mood) based on various information about the user. For example, the information processing apparatus 100 estimates the user's emotion based on the user's behavior information. The information processing apparatus 100 appropriately uses various conventional techniques to estimate the user's emotion. For example, the information processing apparatus 100 estimates the user's emotion using various pieces of input information input by the user, such as content browsed by the user (browsed content) and queries used by the user for searching. For example, the information processing apparatus 100 may estimate the user's emotion by analyzing character information and input information in the browsed content using various conventional techniques such as syntactic analysis as appropriate. For example, the information processing apparatus 100 stores a list of keywords (emotion keyword list) for each emotion in the storage unit 120 (see FIG. 3), and the keyword included in the character information in the browsed content or the input information is the most applicable. The user's emotion may be estimated as the user's emotion. For example, the information processing apparatus 100 uses an emotion keyword list of keywords corresponding to each of positive emotion and negative emotion to identify which of the positive emotion and negative emotion is included in the text information in the browsed content and the input information. The user's emotion may be estimated to be the emotion that has more keywords. Further, the information processing apparatus 100 may estimate the user's emotion using character information obtained by converting the user's utterance and the emotion keyword list.

For example, the information processing apparatus 100 estimates a user's emotion based on sensor information detected by a terminal device used by the user. For example, the information processing apparatus 100 may estimate the user's emotion based on sensor information regarding the user's stride, amplitude, and the like. Further, for example, the information processing apparatus 100 may estimate the user's emotion using biometric information such as the user's heartbeat and blood pressure, image information obtained by imaging the user's face, and the like. Note that the above is just an example, and the information processing apparatus 100 may use any method to estimate the user's emotion, as long as the user's emotion can be estimated. For example, the information processing apparatus 100 may estimate the user's emotion based on the tone of the user's speech. For example, when the tone of the user's speech is high, the information processing apparatus 100 may estimate that the user's emotion is a positive emotion such as joy or pleasure. For example, when the tone of the user's speech is weak, the information processing apparatus 100 may estimate that the user's emotion is a negative emotion such as anger or sadness. For example, the information processing apparatus 100 may estimate that the user's emotion is a positive emotion such as joy or pleasure when the number of products purchased by the user or the amount of money is equal to or greater than a predetermined threshold. For example, the information processing apparatus 100 may estimate that the user's emotion is a negative emotion such as anger or sadness when the number of products purchased by the user or the amount of money is less than a predetermined threshold.

For example, context information may include information indicating whether the temperature corresponds to high, medium, or low. For example, the context information includes information indicating whether the temperature corresponds to high (+5°C or higher), medium (less than ±5°C), or low (-5°C or lower) relative to the average temperature. may be included. For example, the context information may include information indicating whether the atmospheric pressure corresponds to high, medium, or low. For example, the context information may indicate that the atmospheric pressure is high (+1 hectopascal or more), medium (less than ±1 hectopascal), or low (-1 hectopascal or less) with respect to the atmospheric pressure immediately before (for example, 10 minutes or 1 hour ago). ) may be included. For example, the context information may include information regarding the content of the operation. For example, the context information may include various information such as information indicating what kind of operation (push notification), information indicating the category of notification content, and information indicating whether or not it is an app teaser. Note that the above is an example, and the context information may include various types of information without being limited to the above.

For example, the context information may include information about the user's situation estimated based on the situation of the user or the terminal device 10 or the environment (background) of the user or the terminal device 10 . In addition, the context information includes details of content provided to the user, details of content reacted to by the user, attributes of the user, current position of the user, current time, physical environment in which the user is placed, society in which the user is placed, and so on. Various information may be included such as the environment, the user's motion state, and the estimated user's emotion. The context information may also include various types of information such as the state of mind of the user estimated based on the situation of the user or the terminal device 10 or the environment (background) of the user or the terminal device 10 .

The information providing device 50 is an information processing device having various open data. For example, the information providing device 50 provides open data. The information providing device 50 is an information processing device that provides the information processing device 100 with open data, which is data that is permitted to be widely used. For example, the information providing device 50 may be an information processing device that is managed by each local government and provides the data owned by the local government to the outside as open data. Further, for example, the information providing device 50 may be various information processing devices such as a server managed by the Japan Meteorological Agency. The information providing device 50 is not limited to the above, as long as the data (information) is permitted to be used, open data on traffic such as road information, open data on weather such as weather and temperature, geospatial information, disaster prevention / Various open data such as disaster mitigation information, procurement information, and statistical information may be provided.

For example, the information providing device 50 provides open data such as A area and B area to the information processing device 100 . For example, the information providing device 50 provides the information processing device 100 with environment information regarding the A area, the B area, and the like. For example, the information providing device 50 provides the information processing device 100 with various environmental information such as traffic information of the A area. For example, the information providing device 50 provides the information processing device 100 with various environmental information such as traffic information of area B and information on population composition. In this way, the environmental information may include various types of information such as road information and weather information in the vicinity of the user.

Based on the context information about content distribution and the content information about content, the information processing apparatus 100 selects an element of context information (also referred to as a “first element”) related to content distribution and an element of content information (“ It is an information processing device that generates relationship information indicating a combination with a second element. The information processing device 100 is a generation device that generates relationship information indicating a combination of a first element and a second element based on context information and content information. In the example of FIG. 1, the information processing apparatus 100 generates relationship information indicating the combination of the first element and the second element using a model including feature amounts relating to the first element and the second element. The information processing apparatus 100 generates relationship information indicating a combination of the first element and the second element using a model for predicting profitability in content distribution. It should be noted that the profitability of content distribution referred to here may be of various types as long as the distribution of content can bring revenue (profit) to a predetermined entity. Profitability in distribution of content may be profitability in any entity involved in the distribution of content. For example, profitability in distribution of content may indicate the profit earned by the provider of the product when the user purchases the product related to the distributed content. For example, it is estimated that the longer a user browses content, the more likely it is that an action that will generate revenue (profit) for a given entity, such as a user purchasing the content, will occur. may indicate how long the user views the delivered content. That is, the information processing apparatus 100 generates relationship information indicating a combination of the first element and the second element using a model that predicts user behavior with respect to content. Further, the information processing apparatus 100 generates a model for predicting the opening probability of content, such as the model M11 shown in FIG. may be generated.

The information processing apparatus 100 is located in a predetermined area, based on the operation information indicating whether or not the user has performed a predetermined operation on the content notified by push in the terminal device 10 used by the user. This is an information processing apparatus that generates prediction information for predicting whether or not a user performs a predetermined operation on the content when content is pushed to the terminal device 10 used by the user located in the area of . For example, the information processing device 100 acquires various information such as behavior information including position information from the terminal device 10 . Also, for example, the information processing apparatus 100 acquires various information such as open data from the information providing apparatus 50 .

An example of information processing by the information processing apparatus 100 (see FIG. 3) will now be described with reference to FIG. First, the map information MP1 shown in FIG. 1 will be briefly described. Map information MP1 shown in FIG. 1 is a diagram schematically showing an area included in a predetermined range. For example, the map information MP1 is a diagram for schematically showing the status of push notification to the terminal device 10 in the A area and the B area.

In the example of FIG. 1, the information processing apparatus 100 obtains information indicating that predetermined content has been pushed to the terminal device 10 used by a certain user and has been opened by the user (step S1-1). For example, the information processing apparatus 100 sends a push notification of predetermined content (for example, content CT11 or the like) to the terminal device 10 used by the user U1 (see FIG. 4) located in the A area, indicating that the content has been opened by the user U1. Get information. In this way, when "user U* (* is an arbitrary number)" is written, it indicates that the user is identified by the user ID "U*". For example, when "user U1" is described, the user is identified by the user ID "U1".

For example, the information processing apparatus 100 obtains information indicating that predetermined content has been pushed to the terminal device 10 used by the user U1 located in the partial area LC1-1 of the A area, and has been opened by the user U1. . Note that the partial area LC1-1 may be an area within a predetermined range in the A area. The partial area LC1-1 may be one piece of position information representing the range of the partial area LC1-1. For example, the partial area LC1-1 may be position information of the center within the range of the partial area LC1-1. Further, the partial area LC1-1 may be a combination of position information of the center within the range of the partial area LC1-1 and information indicating the range from the center (for example, within 100 m). Further, the partial area LC1-1 may be a list of positions within the range of the partial area LC1-1. Further, in the example of FIG. 1, the information processing apparatus 100 acquires information indicating that the content (content CT11) notified in step S1-1 has been viewed by the user for time RTM1 after opening (step S2 ). For example, the information processing apparatus 100 may acquire information indicating that the content notified in step S1-1 was displayed on the terminal device 10 of the user for the time RTM1 after opening.

Note that the information processing apparatus 100 may use information indicating the viewing time of the user for not only the content distributed to the user, but also the entire content group having a predetermined relationship with the content after the content distribution. For example, the information processing apparatus 100 uses information indicating the viewing time of the user U1 not only for the content CT11 distributed to the user U1, but also for the entire content group having a predetermined relationship with the content CT11 after the content CT11 is distributed. may For example, after distributing the content CT11 to the user U1, the information processing apparatus 100 may use information indicating the browsing time of the user U1 for the entire content group of the site (site STX) to which the content CT11 belongs. That is, the information processing apparatus 100 may generate the prediction information based on the time (profitability) during which the user viewed the content group related to the site STX after the content related to the site STX was distributed to the user. In this case, for example, the information processing apparatus 100 determines the profit (profitability) obtained by the operator who operates the site STX from the user's behavior after the content related to the site STX is distributed to the user. may be

For example, the information processing device 100 stores information acquired from the terminal device 10 in the behavior information storage unit 123 (see FIG. 5). Note that the content may be provided to the terminal device 10 by the information processing device 100 or by another device as long as the information processing device 100 can acquire information used to generate prediction information. . For example, content may be provided to the terminal device 10 by any device as long as the information processing device 100 can acquire information used to generate prediction information.

Further, in the example of FIG. 1, the information processing apparatus 100 acquires information indicating that predetermined content has been pushed to the terminal device 10 used by a certain user and has not been opened by the user (step S1-2). . For example, the information processing apparatus 100 acquires information indicating that a predetermined content was pushed to the terminal device 10 used by the user U2 (see FIG. 4) located in the A area and was not opened by the user U2. For example, the information processing apparatus 100 acquires information indicating that predetermined content has been pushed to the terminal device 10 used by the user U2 located in the partial area LC1-2 of the A area and has not been opened by the user U2. do.

In the example of FIG. 1, to simplify the explanation, the user's operations regarding two push notifications in steps S1-1 and S1-2 are illustrated. It is assumed that a push notification (for example, 10 million times) is sent, and the information processing apparatus 100 acquires action information such as operation information indicating whether or not the user has opened the package in response to the notification. Further, steps S1-1 and S1-2 are for explaining the processing, and either step S1-1 or S1-2 may be performed first, and steps S1-1 and S1-2 may be performed first. may be performed multiple times. Hereinafter, steps S1-1 and S1-2 will be collectively referred to as step S1 when described without distinguishing between them.

Further, the information processing apparatus 100 generates a model used for generating prediction information based on the user's action information regarding the push notification acquired in step S1 (step S3). For example, the information processing apparatus 100 generates a model for predicting the behavior of the user with respect to the distributed content based on the behavior information of the first user in the A area. For example, the information processing apparatus 100 generates a model for predicting the behavior of the user to whom the content was distributed to the content based on the behavior information of the first user in the A area. For example, the information processing apparatus 100 generates a model for predicting user behavior with respect to distributed content based on the first user's context information in the A area. For example, the information processing apparatus 100 generates a model that predicts how much the distributed content will be opened and browsed based on the context information. For example, the information processing apparatus 100 generates a model for predicting profitability in push notifications based on various information stored in the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, and the like. For example, the information processing apparatus 100 creates a model for predicting the behavior of the user with respect to the push-notified content based on various types of information stored in the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, and the like. Generate.

For example, the information processing apparatus 100 generates a model based on the environment information of area A stored in the open data storage unit 121 . For example, the information processing apparatus 100 generates a model based on the user information of the user located in the A area stored in the user information storage unit 122 . For example, the information processing apparatus 100 generates a model based on the behavior information of the user located in the A area stored in the behavior information storage unit 123 . The information processing apparatus 100 generates a model by taking into account the information about the content notified to the user by push. For example, the information processing apparatus 100 generates a model including a feature quantity whose elements are the categories of the content notified to the user by push notification.

Here, the information processing apparatus 100 generates a model including each feature amount corresponding to each element. That is, the information processing apparatus 100 determines the feature amount of the model and learns the weight of each feature amount to generate the model. For example, the information processing apparatus 100 may generate a model using feature amounts obtained by combining elements of context information by various conventional techniques such as the Kronecker product. For example, the information processing apparatus 100 generates a model including feature amounts that combine two or more elements. For example, the information processing apparatus 100 determines a feature amount that combines elements of context information. For example, the information processing apparatus 100 determines a feature amount that combines elements between categories of context information.

For example, the information processing apparatus 100 can create a model based on each combination of the element group EL1 corresponding to the context related to content distribution and the element group EL2 corresponding to the details of the content such as the content category among the context information. Determine features. For example, the information processing apparatus 100 determines each combination of the first element of the element group EL1 and the second element of the element group EL2 as the feature amount of the model. For example, the information processing apparatus 100 identifies each combination of the first element of the element group EL1 such as “Cafe A” and the second element of the element group EL2 such as “soccer” and “entertainment” as the characteristics of the model. Decide on quantity. Cafe A may be a cafe of a specific chain store or franchise store, or may be a store belonging to the category of cafes such as "sports cafe", "dining cafe", "cat cafe", etc. It may be a cafe belonging to a category (type).

In the example of FIG. 1, the information processing device 100 corresponds to the feature amount #1 corresponding to the combination of Cafe A and soccer, the feature amount #2 corresponding to the combination of Cafe A and entertainment, and the combination of Cafe A and economy. The feature amount #3 and the like to be used are determined as the feature amounts of the model M1. Although illustration is omitted in the example of FIG. and baseball, a feature #12 corresponding to a combination of cafe A and tennis, a feature #13 corresponding to a combination of cafe A and politics, and the like. The feature amount of M1 is determined.

Further, the feature amount included in the model is not limited to a combination of elements, and the information processing apparatus 100 may generate a model including a feature amount corresponding to only one element of the context information. The information processing apparatus 100 may generate a model including feature amounts in which three or more various elements are combined. For example, the information processing apparatus 100 may generate a model including a feature amount corresponding only to the element "cafe A" or a feature amount corresponding only to the element "soccer". The combination of the elements described above is an example, and various combinations such as a combination of elements in the element group EL1 may be used. For example, the information processing apparatus 100 stores context element list information of context-related elements such as “cafe A” and “morning” stored in the storage unit 120 (see FIG. 3), “soccer”, “entertainment”, and “economy”. , "baseball", etc., may be used to determine the combination. For example, the information processing apparatus 100 may generate information indicating each combination by combining a first element selected from the context element list information and a second element selected from the content element list information. For example, the information processing apparatus 100 may determine tens of millions or hundreds of millions of feature amounts as feature amounts to be included in the model M1.

For example, the information processing apparatus 100 may generate a model for predicting the behavior of the user with respect to the push-notified content, using information when the push-notified content is opened as correct information (positive example). For example, the information processing apparatus 100 uses, as correct information, various context information such as user information of the user, location information of the user, behavior information of the user, and environment information of the user when the content notified by the push notification is opened. A model may be generated that predicts user behavior with respect to the notified content. For example, when the user U1 opens the content notified by the push notification in the partial area LC1-1, the information processing apparatus 100 stores the user information of the user U1, the location information of the user U1, the behavior information of the user U1, and the environment information of the user U1. A model for predicting the behavior of the user with respect to the push-notified content may be generated using various context information such as the correct answer information. For example, the information processing apparatus 100 generates the model M1 using context information regarding distribution of content to the user U1. For example, the information processing apparatus 100 generates the model M1 using information regarding content distributed to the user U1. For example, the information processing apparatus 100 generates the model M1 using information regarding the content CT11 distributed to the user U1. For example, the information processing apparatus 100 generates the model M1 using information indicating the category of content distributed to the user U1. For example, information processing apparatus 100 generates model M1 using content information indicating that the category of content distributed to user U1 is "soccer".

In the example of FIG. 1 , when the content is opened, the information processing apparatus 100 selects one of a plurality of levels indicating the degree of user behavior regarding profitability based on the user's viewing time of the content after opening the content. Whichever is applicable determines the level of user activity in each delivery.

For example, when the user browses content for less than 5 seconds, the information processing apparatus 100 learns to output a score of "1" with the user's action level in the distribution being level LV1. Generate a model that For example, when the time RTM1, which is the browsing time of the user U1 shown in FIG. If so, a model that predicts is generated by learning to output a score of "1".

Further, for example, the information processing apparatus 100 learns to output a score of "2" with the level LV2 of the user's behavior in the distribution when the viewing time of the user's content is 5 seconds or more and less than 1 minute. to generate a predictive model. For example, when the time RTM1, which is the browsing time of user U1 shown in FIG. Generates a model that predicts by learning to output a score of "2" when input.

Further, for example, the information processing apparatus 100 learns to output a score of "3" with the level LV3 of the user's behavior in the distribution when the viewing time of the content by the user is 1 minute or more and less than 5 minutes. to generate a predictive model. For example, when the time RTM1, which is the browsing time of the user U1 shown in FIG. A model may be generated that predicts by learning to output a score of "3" when input. Note that the above is just an example, and the information processing apparatus 100 may use various levels, and the information processing apparatus 100 may generate models for predicting various types of profitability, not limited to viewing time.

In addition, for example, the information processing apparatus 100 generates a model for predicting the behavior of the user with respect to the push-notified content, using information when the push-notified content is not opened as incorrect information (negative example). good too. For example, when the user does not open the content, the information processing apparatus 100 learns to output a score of "0" with the level of the user's behavior in the distribution being the level LV0. to generate For example, the information processing apparatus 100 may generate a predictive model by learning to output a score of “0” when context information is input when the push-notified content has not been opened. good.

For example, the information processing apparatus 100 uses various context information such as user information of the user, location information of the user, behavior information of the user, and environment information of the user when the push-notified content is not opened as incorrect answer information. , may generate a model that predicts user behavior with respect to pushed content. For example, if the user U2 does not open the content of which the push notification is sent in the partial area LC1-2, the information processing apparatus 100 sets the user information of the user U2, the location information of the user U2, the behavior information of the user U2, and the content of the user U2. Various context information such as environmental information may be used as incorrect answer information to generate a model for predicting the behavior of the user with respect to the content notified by the push notification. For example, the information processing apparatus 100 generates the model M1 using context information regarding the distribution of content to the user U2. For example, the information processing device 100 generates the model M1 using information about the content distributed to the user U2. For example, the information processing apparatus 100 generates the model M1 using information regarding the content CT11 distributed to the user U2. For example, the information processing device 100 generates the model M1 using information indicating the category of content distributed to the user U2. For example, the information processing apparatus 100 generates the model M1 using content information indicating that the category of the content delivered to the user U2 is "soccer".

In the example of FIG. 1, the information processing apparatus 100 generates model information of the model M1 as shown in the model information MD. For example, the information processing apparatus 100 generates a model M1 as shown in the model information storage unit 125 (see FIG. 7).

The model M1 shown in the model information storage unit 125 in FIG. Each weight (value) is associated and stored. For example, in the example shown in FIG. 7, the model information about the model M1 is such that the weight of the feature value #1 corresponding to the combination of cafe A and soccer is "0.5", and the feature value #1 corresponding to the combination of cafe A and performing arts is "0.5". 2 is "-0.1", the weight of feature #3 corresponding to the combination of cafe A and economy is "0.2", and so on. Note that each model may be generated using various conventional techniques related to machine learning as appropriate. For example, a model may be generated using a technology related to supervised learning machine learning such as SVM (Support Vector Machine). Note that the above description of model generation is an example, and model generation may be performed by a learning method appropriately selected according to obtainable information and the like. In this manner, the information processing apparatus 100 generates the model M1 by learning the weight of each feature amount such as the feature amount #1 to the feature amount #3 of the model M1.

For example, the information processing apparatus 100 receives as input context information regarding content distribution, and generates a model M1 that outputs a score based on the input. For example, the information processing apparatus 100 receives context information as input and generates a model M1 that outputs a score indicating the possibility of actions by the user with respect to the distributed content corresponding to the input context information. For example, the information processing apparatus 100 receives context information as input and generates a model M1 that outputs a score indicating the level of user behavior (profitability) in distributing content corresponding to the input context information. Note that the information processing apparatus 100 may generate, as the model M1, a model that outputs a score indicating the probability that the user to whom the content is distributed performs the opening operation, as shown in the model M11 of FIG. 11 .

Then, the information processing apparatus 100 generates relationship information indicating the combination of the first element and the second element based on the generated model M1 (step S4). In the example of FIG. 1, the information processing apparatus 100 generates relationship information indicating the correspondence between the combination of the first element and the second element and the degree of relationship between the user's opening behavior for the content, based on the generated model M1. Generate. Based on the generated model M1, the information processing apparatus 100 generates relationship information indicating the degree of relationship between the first element corresponding to the context related to content distribution and the second element corresponding to the content, and the browsing behavior of the user for the content. Generate.

For example, the information processing apparatus 100 generates relationship information indicating the degree of relationship between the content and the user's predetermined behavior based on the weight of each feature amount of the model M1. The information processing apparatus 100 generates relationship information by selecting a feature amount whose corresponding weight satisfies a predetermined condition from among the feature amounts of the model M1. The information processing apparatus 100 generates relationship information by selecting a feature amount whose absolute value of the corresponding weight is larger than a predetermined threshold among the feature amounts of the model M1. In the example of FIG. 1, the information processing apparatus 100 generates relational information by selecting, from among the feature amounts of the model M1, the feature amount whose corresponding weight absolute value is greater than the threshold "0.3". Note that the threshold is not limited to "0.3" and may be set as appropriate. Further, for example, the information processing apparatus 100 generates relational information in which relational information corresponding to each feature quantity is ranked so that a feature quantity having a large absolute value of weight among the feature quantities of the model M1 is ranked high. Generate. The information processing apparatus 100 then stores the generated relationship information in the relationship information storage unit 127 .

"Target behavior" in the relationship information storage unit 127 in FIG. 1 indicates a predetermined behavior of the target user. "Ranking" in the relational information storage unit 127 in FIG. 1 indicates the order (ranking) of the degree of relation between each piece of relational information and the target action. “Relationship information ID” in the relation information storage unit 127 in FIG. 1 indicates information for identifying relation information. "Weight" in the relationship information storage unit 127 in FIG. 1 indicates the weight of the feature quantity corresponding to the relationship information. "Element #1" to "Element #3" etc. in the relationship information storage unit 127 in FIG. 1 indicate the elements of the feature quantity corresponding to the relationship information.

The relational information storage unit 127 in FIG. 1 indicates that the target action is the user's action on the content. For example, it indicates that the target action is a user's action such as opening a content. In addition, the relationship information (relationship information ST1) identified by the relationship information ID "ST1" corresponds to the feature amount with the weight of "0.5", the element #1 is "Cafe A", and the element #2 is Indicates that it is "soccer". That is, the relationship information ST1 corresponds to the feature amount #1 of the model M1.

As described above, in the example of FIG. 1, the information processing apparatus 100 ranks the relationship information ST1 corresponding to the feature quantity #1 because the feature quantity #1 has the largest absolute value of the weight among the feature quantities of the model M1. is determined as first place. The information processing apparatus 100 generates relationship information indicating that the combination of the elements “Cafe A” and “soccer” of the feature amount #1 among the feature amounts of the model M1 has a relationship with the user's actions such as opening contents. do. Since the weight of the relational information ST1 of the combination of the elements “cafe A” and “soccer” is a positive value, the information processing apparatus 100 determines that when the location is “cafe A” and the content category is “soccer”, It can be estimated that the user is likely to take action on the content.

In addition, since the absolute value of the weight of the feature amount corresponding to the combination of the elements “morning” and “entertainment” among the feature amounts of the model M1 is the second largest after the feature amount #1, the information processing apparatus 100 determines that the feature amount , is determined to be the second rank. Since the weight of the relational information ST2 of the combination of the elements "morning" and "entertainment" is a negative value, the information processing apparatus 100 determines that the time is "morning" and the content category is "entertainment". It can be estimated that the likelihood of taking action on the content is low.

Note that the information processing apparatus 100 may generate the relationship information by selecting a feature amount whose weight is greater than the threshold instead of the absolute value of the weight of the feature amount. In this case, the information processing apparatus 100 generates relationship information indicating that the element of the feature quantity with a large weight influences the user's action such as opening the content.

Then, the information processing apparatus 100 provides content to the user based on the generated relationship information. First, the information processing apparatus 100 determines a user (target user) to whom content is to be provided (step S5). In the example of FIG. 1, the information processing apparatus 100 determines the user U3 located at the shop SP1, which is the cafe A in the B area, as the target user. The information processing apparatus 100 determines the user U3 who is the second user located in the B area, which is an area other than the A area, as the target user.

Then, the information processing apparatus 100 determines content to be provided to the user U3, who is the target user, based on the generated relationship information. The information processing apparatus 100 selects, from among the relational information stored in the relational information storage unit 127, the relational information whose place-related element is “Cafe A”, “Cafe”, or the like. For example, the information processing apparatus 100 selects a predetermined number (for example, 5 or 10) of the relational information having the location-related elements such as "cafe A" and "cafe" in descending order of rank. In the example of FIG. 1, the information processing apparatus 100 selects the relational information ST1, the relational information ST4, etc., including the element "Cafe A" in descending order of rank.

Here, since the weight of the selected relationship information ST1 is a positive value, the information processing apparatus 100 distributes the content of the category “soccer” to the user located at the target location “Cafe A”. Assume that it is likely to be opened. In addition, since the weight of the selected relational information ST4 is a negative value, the information processing apparatus 100 may open the content when the content of the category “baseball” is distributed to the user located in the place “cafe A”. presumed to be low. Therefore, the information processing apparatus 100 determines to provide soccer-related content (soccer-related content CT31) to the user U3 located at the place “Cafe A”.

The information processing device 100 then provides the soccer-related content CT31 to the user U3 (step S6). The information processing device 100 provides the soccer-related content CT31 to the terminal device 10-3 used by the user U3. The information processing device 100 transmits the soccer-related content CT31 to the terminal device 10-3 used by the user U3. The information processing device 100 distributes the soccer-related content CT31 to the terminal device 10-3 used by the user U3.

As described above, by using the generated model M1, the information processing apparatus 100 can appropriately generate the information indicating the relationship between the content distribution context and the content. In the example of FIG. 1, the information processing apparatus 100 uses the relationship information generated using the model M1 generated from the information of the first user in the A area to identify the second user in the B area, which is different from the A area. Appropriate content can be provided to user U3.

[1-1. Example of generating information]
Note that FIG. 1 is an example, and the information processing apparatus 100 may generate various types of information without being limited to the above. This point will be described below with reference to FIG. FIG. 11 is a diagram illustrating an example of information processing according to the embodiment; In FIG. 11, the description of the same points as in FIG. 1 may be omitted as appropriate. In the example of FIG. 11, the information processing apparatus 100 (see FIG. 3) pushes to another area AR21 (area B) based on action information such as user operation information related to push notification in a predetermined area AR11 (area A). It shows the case of generating prediction information about notifications. In the example of FIG. 11, push notification is shown as one form of content distribution, but content distribution may be in various forms. For example, content distribution in the information processing system 1 is not limited to push notification, and may be so-called pull-type content distribution in which content is requested from the terminal device 10 .

For example, in the example of FIG. 11, due to a lack of action information such as actual user operation information related to push notifications in area B, distribution information of push notification opening rates in area B is generated with high accuracy. indicate when it is difficult. For example, the push notification may be information output from the terminal device 10 regardless of the operation of the terminal device 10 by the user. For example, push notification may be to actively notify the user of predetermined information. Further, the opening rate referred to here indicates, for example, the rate at which the user has performed the opening operation for the push notification displayed on the terminal device 10 . Also, the opening operation may be a user's predetermined operation such as selection or designation for the push notification. For example, the unsealing operation may be a predetermined operation by the user that causes the terminal device 10 to display information about push notifications. For example, in the example of FIG. 11, as shown in step S10, it is assumed that the number of notifications in area B is small. It should be noted that step S10 is for illustrating that the number of notifications for the B area is small, and may not be included in the processing procedure.

First, the map information MP1-1 to MP1-3 shown in FIG. 11 will be briefly described. The map information MP1-1 to MP1-3 shown in FIG. 11 are diagrams schematically showing areas included in a predetermined range. For example, the map information MP1-1 is a diagram for schematically showing the state of push notification to the terminal device 10 in the A area and the B area. Further, for example, the map information MP1-2 is a diagram for schematically showing the distribution information of the open rate of the push notification to the terminal device 10 in the A area. Further, for example, the map information MP1-3 is a diagram for schematically showing prediction information of the opening rate of the push notification to the terminal device 10 in the B area. Also, the ranges (areas) shown in the map information MP1-1 to MP1-3 are the same area. Further, hereinafter, the map information MP1-1 to MP1-3 will be referred to as map information MP1 when they are not distinguished from each other.

In the example of FIG. 11, the information processing apparatus 100 obtains information indicating that predetermined content has been pushed to the terminal device 10 used by a certain user and has been opened by the user (step S11-1). For example, the information processing apparatus 100 sends a push notification of predetermined content (for example, content CT11 or the like) to the terminal device 10 used by the user U1 (see FIG. 4) located in the A area, indicating that the content has been opened by the user U1. Get information. For example, the information processing apparatus 100 acquires information indicating that predetermined content has been pushed to the terminal device 10 used by the user U1 located in the partial area LC1-1 of the A area, and has been opened by the user U1. . Note that the partial area LC1-1 may be an area within a predetermined range in the A area. Further, in the example of FIG. 11, the information processing apparatus 100 may acquire information indicating that the content (content CT11) notified in step S11-1 has been viewed by the user for time RTM1 after opening. . For example, the information processing apparatus 100 may acquire information indicating that the content notified in step S11-1 was displayed on the terminal device 10 of the user for the time RTM1 after opening.

For example, the information processing device 100 stores information acquired from the terminal device 10 in the behavior information storage unit 123 (see FIG. 5). Note that the content may be provided to the terminal device 10 by the information processing device 100 or by another device as long as the information processing device 100 can acquire information used to generate prediction information. . For example, content may be provided to the terminal device 10 by any device as long as the information processing device 100 can acquire information used to generate prediction information.

Further, in the example of FIG. 11, the information processing apparatus 100 acquires information indicating that predetermined content has been pushed to the terminal device 10 used by a certain user and has not been opened by the user (step S11-2). . For example, the information processing apparatus 100 obtains information indicating that predetermined content was pushed to the terminal device 10 used by the user U2 (see FIG. 4) located in the A area and was not opened by the user U2. For example, the information processing apparatus 100 acquires information indicating that predetermined content was pushed to the terminal device 10 used by the user U2 located in the partial area LC1-2 of the A area, and was not opened by the user U2. do.

In the example of FIG. 11, to simplify the explanation, the user's operation regarding two push notifications in steps S11-1 and S11-2 is illustrated. It is assumed that a push notification (for example, 10 million times) is sent, and the information processing apparatus 100 acquires action information such as operation information indicating whether or not the user has opened the package in response to the notification. Further, steps S11-1 and S11-2 are for explaining the processing, and either step S11-1 or S11-2 may be performed first, and steps S11-1 and S11-2 may be performed first. may be performed multiple times. Hereinafter, steps S11-1 and S11-2 will be collectively referred to as step S11 when described without distinguishing between them.

Then, the information processing apparatus 100 generates distribution information based on action information such as operation information related to push notification (step S12). For example, the information processing apparatus 100 generates the distribution information by calculating the opening rate of the content based on the operation information regarding the push notification. For example, the information processing apparatus 100 stores the distribution information generated based on the operation information acquired in step S11 in the distribution information storage unit 124 (see FIG. 6). For example, the information processing apparatus 100 stores distribution information regarding the A area as the aggregation target area in the distribution information storage unit 124 .

The "aggregation target area" shown in the distribution information storage unit 124 in FIG. 6 indicates an area from which distribution information is collected. Also, "distribution information" shown in the distribution information storage unit 124 in FIG. 6 indicates distribution information collected about the aggregation target area. A "partial area" shown in the distribution information storage unit 124 in FIG. 6 indicates a partial area within the aggregation target area. The "number of pushes" shown in the distribution information storage unit 124 in FIG. 6 indicates the number of push notifications made to users located in the corresponding partial area. Further, the "number of times of opening" shown in the distribution information storage unit 124 in FIG. 6 indicates the number of times the user located in the corresponding partial area has opened the push notification. The "opening rate (%)" indicates the percentage of push notifications opened by users located in the corresponding partial area (=number of times of opening/number of times of pushing×100). For example, the "opening rate (%)" is a value between 0 and 100 (%). For example, the information processing apparatus 100 calculates the opening rate.

For example, in the example shown in the distribution information storage unit 124 in FIG. Indicates that a push notification was made. Also, for example, it indicates that the user located in the partial area LC1-1 has opened the push notification "2000 (times)". Also, for example, the opening rate of users located in the partial area LC1-1 is "20(%)" because 2,000 out of 10,000 notifications have been opened. For example, the information processing apparatus 100 calculates the opening rate of users located in the partial area LC1-1 as "20% (=2000/10000*100)".

Note that the information processing apparatus 100 may calculate the opening rate based on various types of information, not limited to the above. For example, the information processing apparatus 100 may use the score itself output by the model M11 to calculate the opening rate of users located in the partial area LC1-1. For example, the information processing apparatus 100 calculates the average value of the scores output by the model M11 to which the context information of the content distribution in the partial area LC1-1 is input as the open rate of the users located in the partial area LC1-1. good too. For example, the information processing apparatus 100 divides the total score output by the model M11 for content distribution in the partial area LC1-1 by the number of outputs, thereby calculating the open rate of users located in the partial area LC1-1. You may For example, information processing apparatus 100 assumes that the total value of scores output by model M11 for content distribution in partial area LC1-1 is “45 (=0.2+0.5+ . . . )” and the number of outputs is “100”. In some cases, the opening rate of users located in the partial area LC1-1 may be calculated as "45% (=45/100*100)".

In the example of FIG. 11, the information processing apparatus 100 generates distribution information DM11 with the A area as the aggregation target area. The distribution information DM11 shown in the A area of the map information MP1-2 in FIG. 11 shows the degree of opening rate in each partial area in a display form like a heat map. In the example of FIG. 11, a darker hatched area indicates a higher opening rate. In the example of FIG. 11, the distribution information DM11 indicates that the partial area LC1-1 in the A area has a high opening rate. Also, in the example of FIG. 11, the distribution information DM11 indicates that the partial area LC1-2 in the A area has a low opening rate. That is, in area A, the opening rate is high near the center and decreases as the distance from the center increases.

Further, the information processing apparatus 100 generates a model used for generating prediction information based on the action information such as the operation information related to the push notification acquired in step S11 (step S13). For example, the information processing apparatus 100 generates a model for predicting whether to open the push notification based on action information such as operation information in the A area. For example, the information processing apparatus 100 generates a model for predicting user behavior (profitability) in content distribution based on context information. For example, the information processing apparatus 100 generates a model for predicting whether distributed content will be opened based on context information. For example, the information processing apparatus 100 generates a model for predicting whether to open the push notification based on various information stored in the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, and the like. . For example, the information processing apparatus 100 generates a model based on the environment information of area A stored in the open data storage unit 121 . For example, the information processing apparatus 100 generates a model based on the user information of the user located in the A area stored in the user information storage unit 122 . For example, the information processing apparatus 100 generates a model based on the behavior information of the user located in the A area stored in the behavior information storage unit 123 . Further, for example, the information processing apparatus 100 may generate a model by taking into account information related to content that has been pushed to the user.

For example, the information processing apparatus 100 may generate a model for predicting whether or not to open the push notification, using information when the push-notified content is opened as correct information (positive example). For example, the information processing apparatus 100 may generate the model M11 by learning to output a score of "1" when context information is input when the push-notified content is opened. For example, the information processing apparatus 100 uses, as correct information, various context information such as user information of the user, location information of the user, behavior information of the user, and environment information of the user when the content notified by the push notification is opened. A model may be generated that predicts whether a notification will be opened. For example, when the user U1 opens the content notified by the push notification in the partial area LC1-1, the information processing apparatus 100 stores the user information of the user U1, the position information of the user U1, the behavior information of the user U1, and the environment information of the user U1. A model for predicting whether or not to open a push notification may be generated using various context information such as the correct answer information.

Further, for example, the information processing apparatus 100 may generate a model for predicting whether or not to open the push notification, using information when the content notified by the push notification is not opened as incorrect information (negative example). . For example, the information processing apparatus 100 may generate the model M11 by learning to output a score of "0" when context information is input when the push-notified content has not been opened. . For example, the information processing apparatus 100 uses various context information such as user information of the user, location information of the user, behavior information of the user, and environment information of the user when the push-notified content is not opened as incorrect answer information. , may generate a model that predicts whether to open a push notification. For example, when the user U2 does not open the content notified by the push notification in the partial area LC1-2, the information processing apparatus 100 sets the user information of the user U2, the location information of the user U2, the behavior information of the user U2, the Various context information such as environment information may be used as incorrect information to generate a model for predicting whether or not to open the push notification.

In the example of FIG. 11, the information processing apparatus 100 generates model information of the model M11 as shown in the model information MD. For example, the information processing apparatus 100 generates a model M11 as shown in the model information storage unit 125 (see FIG. 7). For example, the information processing apparatus 100 receives as input context information regarding content distribution, and generates a model M11 that outputs a score based on the input. For example, the information processing apparatus 100 receives context information as input and generates a model M11 that outputs a score indicating the probability that a user will perform a predetermined action in distributing content corresponding to the input context information. In this case, the information processing apparatus 100 may use the score output by the model M11 as a value indicating the probability that the user will perform a predetermined action. For example, the information processing apparatus 100 may use the score output by the model M11 as a value indicating the probability that the user will open the content. For example, the information processing apparatus 100 receives context information as an input and generates a model M11 that outputs a score indicating the degree of possibility that a user will perform a predetermined action in distributing content corresponding to the input context information. For example, the information processing apparatus 100 receives context information as input, and generates a model M11 that outputs a score indicating the degree (level) of a user performing a predetermined action in distributing content corresponding to the input context information.

For example, when the score output by the model M11 to which context information such as user information and environment information around the user's location is input is "0.7", the information processing apparatus 100 detects that the user has opened the push notification. It may be determined that the probability of doing so is "70% (=0.7*100)". For example, the information processing apparatus 100 may associate scores in the range of 0 to 1 output by the model M11 with 0% to 100%. For example, when the score output by the model M11 to which the context information is input is "0", the information processing apparatus 100 has a probability of "0% (=0*100)" that the user will open the push notification. can be determined. For example, when the score output by the model M11 to which the context information is input is "1", the information processing apparatus 100 has a probability of "100% (=1*100)" that the user will open the push notification. can be determined.

Further, for example, when the model M11 outputs a value less than 0, that is, a negative score, the information processing apparatus 100 may associate the negative score with 0%. For example, when the score output by the model M11 to which the context information is input is "-1", the information processing apparatus 100 determines that the probability of the user opening the push notification is "0%". good. In this case, for example, the information processing apparatus 100 generates the model M11 by learning to output a score of 0 or less when context information is input when the push-notified content has not been opened. may

Further, for example, when the model M11 outputs a score greater than 1, the information processing apparatus 100 may associate the score greater than 1 with 100%. For example, when the score output by the model M11 to which the context information is input is "3", the information processing apparatus 100 may determine that the probability that the user will open the push notification is "100%". . In this way, the information processing apparatus 100 receives context information as input and generates a model M11 that outputs a score indicating the probability (degree) of a user performing a predetermined action in the distribution of content corresponding to the input context information. good too. In this case, for example, the information processing apparatus 100 generates the model M11 by learning to output a score of 1 or more when the context information when the push-notified content is opened is input. good too. For example, the information processing apparatus 100 may generate the model M11 by learning to output a score of 1 or more when context information is input according to revenue from opened content.

For example, the information processing apparatus 100 receives as input context information such as user information and environmental information around the user's position, and generates a model M11 that outputs a score based on the input. For example, the information processing apparatus 100 receives context information and generates a model M11 that outputs a score based on the input. For example, the information processing apparatus 100 receives as input context information such as user information and environment information around the user's position, and based on the input, the score output by the model M11 is a predetermined threshold value (for example, "0.5 , etc.), it may be determined that the user opens the content of which the push notification is given. For example, the information processing apparatus 100 may store the generated model M11 in the model information storage unit 125 (see FIG. 7). Note that the model M11 shown in FIG. 11 may be the same model as the model M1 shown in FIG. Note that each model may be generated using various conventional techniques related to machine learning as appropriate. For example, a model may be generated using a technology related to supervised learning machine learning such as SVM (Support Vector Machine).

Then, the information processing device 100 generates prediction information based on the generated model M11 (step S14). For example, the information processing apparatus 100 generates prediction information by calculating the content opening rate based on the generated model M11. For example, the information processing apparatus 100 generates prediction information regarding the B area based on the model M11 and various types of information regarding the B area, which is the prediction target area.

The information processing apparatus 100 acquires open data of the prediction target area. In the example of FIG. 11, the information processing apparatus 100 acquires the environment information ED of the B area. For example, the information processing apparatus 100 may acquire various types of information such as environment information of area B from the open data storage unit 121 (see FIG. 3).

Also, for example, the information processing apparatus 100 acquires user information located in the prediction target area. In the example of FIG. 11, the information processing apparatus 100 acquires the user information UD of the B area. For example, the information processing apparatus 100 acquires user information such as that shown in the user information storage unit 122 (see FIG. 4) and user behavior information such as that shown in the behavior information storage unit 123 (see FIG. 5).

For example, the information processing apparatus 100 may generate prediction information for the B area for a user whose position information stored in the user information storage unit 122 (see FIG. 4) corresponds to the B area. For example, the information processing apparatus 100 generates prediction information of the B area for the user U3 whose position information is in the partial area LC2-1 in the B area, the user U4 whose position information is in the partial area LC2-2 in the B area, and the like. You may

For example, the information processing apparatus 100 may predict whether the user U3 will open the push notification by inputting context information such as the user information of the user U3 and the environment information of the user U3 into the model M11. For example, the information processing apparatus 100 inputs various information including position information and behavior information of the user U3 to the model M11, and when the model M11 outputs a score less than a predetermined threshold, the user U3 opens the push notification. You can decide not to. In this case, the information processing apparatus 100 stores information indicating that the user U3 does not open the push notification in the storage unit 120 (see FIG. 3), or notifies that the push notification is performed once in the partial area LC2-1. The information shown may be stored in the prediction information storage unit 126 (see FIG. 8).

Further, for example, the information processing apparatus 100 may predict whether the user U4 will open the push notification by inputting context information such as the user information of the user U4 and the environment information of the user U4 to the model M11. For example, the information processing apparatus 100 inputs various information including position information and behavior information of the user U4 to the model M11, and when the model M11 outputs a score equal to or higher than a predetermined threshold, the user U4 opens the push notification. Then you can judge. In this case, the information processing apparatus 100 stores information indicating that the push notification is opened by the user U4 in the storage unit 120, or indicates that the push notification is sent once in the partial area LC2-2 and is opened. Information may be stored in the prediction information storage unit 126 (see FIG. 8).

For example, the information processing apparatus 100 may generate prediction information for users located in area B during a predetermined time period. For example, the information processing apparatus 100 may generate prediction information corresponding to area B in the morning for users located in area B in the morning time period (6:00 to 10:00). For example, the information processing apparatus 100 may generate prediction information corresponding to the B area at night for users located in the B area during the night time period (19:00 to 23:00).

As described above, the information processing apparatus 100 generates prediction information regarding the B area based on the model M11 and various information. For example, the information processing apparatus 100 stores prediction information regarding the B area as a prediction target area in the prediction information storage unit 126 .

“Prediction target area” shown in the prediction information storage unit 126 in FIG. 8 indicates an area for which prediction information is generated. "Prediction information" shown in the prediction information storage unit 126 in FIG. 8 indicates prediction information generated for the prediction target area. A “partial area” shown in the prediction information storage unit 126 in FIG. 8 indicates a partial area within the prediction target area. The “number of pushes” shown in the prediction information storage unit 126 in FIG. 8 indicates the number of times that a push notification is predicted to be given to the user located in the corresponding partial area in the generation of prediction information. Further, the "opening count" shown in the prediction information storage unit 126 in FIG. 8 indicates the number of times it is predicted that the push notification will be opened by the user located in the corresponding partial area in the generation of the prediction information. Also, the "opening rate (%)" shown in the prediction information storage unit 126 in FIG. /number of pushes×100). For example, the "opening rate (%)" is a value between 0 and 100 (%). For example, the information processing apparatus 100 calculates a user's opening rate.

For example, in the example shown in the prediction information storage unit 126 in FIG. Indicates that a push notification is expected. Also, for example, it indicates that the user located in the partial area LC2-1 in the B area is expected to open the push notification "40 (times)". Further, for example, the opening rate of users located in the partial area LC2-1 in the B area is "2 (%)" because it is predicted that 40 out of 2,000 notifications will be opened. indicates For example, the information processing apparatus 100 calculates the opening rate of users located in the partial area LC2-1 as "2% (=40/2000*100)".

In the example of FIG. 11, the information processing apparatus 100 generates prediction information PM11 with B area as the prediction target area. The prediction information PM11 shown in the B area of the map information MP1-3 in FIG. 11 shows the degree of opening rate in each partial area in a display form like a heat map. In the example of FIG. 11, the darker the shaded area, the higher the predicted opening rate. In the example of FIG. 11, the prediction information PM11 indicates that the lower right partial area in the B area has a high opening rate. In the example of FIG. 11, the prediction information PM11 indicates that the opening rate is slightly high in the partial area LC2-2 in the B area. In addition, in the example of FIG. 11, the prediction information PM11 indicates that the opening rate is low in the partial area LC2-1 and the like in the B area. That is, in the B area, the opening rate is high near the lower right portion, and the opening rate decreases as the distance from the lower right portion increases.

In this way, the information processing apparatus 100 generates prediction information for predicting the open rate of push notifications for area B where the track record of push notifications is insufficient. As a result, the information processing apparatus 100 can accurately perform various services related to push notification even in area B, where the push notification performance is insufficient. For example, the information processing apparatus 100 can generate distribution information using information about a point where information can be easily collected, such as an area with a large population. On the other hand, it is difficult for the information processing apparatus 100 to generate distribution information using the information of an area where the amount of information that can be acquired is insufficient due to reasons such as low population. Therefore, the information processing apparatus 100 uses the information of the area with sufficient information to generate the prediction information of the area with insufficient information, thereby enabling desired prediction for the area with insufficient information. In this way, the information processing apparatus 100 makes it possible to predict the presence or absence of the user's predetermined operation in other obtainable areas with an insufficient track record based on the information of the predetermined area where the track record is sufficiently obtainable. be able to. Also, for example, the information processing apparatus 100 can generate prediction information indicating that the opening rate tends to be high in the vicinity where a specific POI is located. For example, the information processing apparatus 100 can generate prediction information or the like indicating that an area with many users of a specific attribute tends to have a high opening rate. Further, for example, the information processing apparatus 100 generates prediction information according to the time zone even at the same position, thereby generating prediction information indicating that the opening rate increases at the timing when a predetermined event is held. can be generated.

Various predictions can be made based on the prediction information generated by the information processing apparatus 100 as described above. For example, when the information processing apparatus 100 generates prediction information based on location (area, etc.) and time (time period, etc.), the generated prediction information makes it possible to predict the timing at which the opening rate increases in a certain area. . For example, when the information processing device 100 generates prediction information based on the location (area, etc.) and time (time zone, etc.), the generated prediction information indicates the area near the bus stop and the timing at which students gather. Therefore, it is possible to predict that the opening rate near the bus stop will be high. In addition, the information processing apparatus 100 generates prediction information based on the position (area, etc.) and time (time zone, etc.). Moreover, it is possible to predict that the opening rate will be high near the bus stop just before the bus arrives.

Further, for example, when the information processing device 100 generates prediction information based on the position (area, etc.) and time (time zone, etc.), the generated prediction information can be used to determine the area near the racetrack where an important race will be held. In addition, at the timing when an important race starts, it is possible to predict that the opening rate near the racecourse will be high. Further, for example, when the information processing device 100 generates prediction information based on the position (area, etc.) and time (time zone, etc.), the generated prediction information can be used to determine the area near the racetrack where an important race will be held. Moreover, at a predetermined timing (for example, 20 minutes before the start of an important race) before the start of the race, it is possible to predict that the opening rate near the racecourse will be high. Further, the information processing apparatus 100 may use such prediction information to charge a higher fee for advertisement distribution in an area during a time period in which the opening rate in the area is higher than in other time periods. Further, for example, the information processing apparatus 100 uses such prediction information to determine the display unit price for the advertisement frame in the content distributed to the terminal apparatus 10 located in the area during the time period when the opening rate in the area is high. may be higher than other time periods. Note that the above is just an example, and the information processing apparatus 100 may use the prediction information for various services.

For example, the information processing apparatus 100 may perform push notification based on the generated prediction information. For example, the information processing apparatus 100 may send a push notification to a user positioned in a heavily hatched portion (lower right in area B) in the prediction information PM11. Further, for example, the information processing device 100 may provide the generated prediction information to an external information processing device. For example, the information processing device 100 may provide prediction information to an external device that provides a content providing service.

[1-1-1. Generation of prediction information by simulation]
As described above, the information processing apparatus 100 generates the prediction information based on the model M11 and the context information such as the behavior information of the user located in the B area, which is the prediction target area, and the surrounding environment information. However, the prediction information may be generated by a predetermined simulation. In this case, the storage unit 120 of the information processing apparatus 100 may include a simulator information storage unit (not shown) that stores various information related to a predetermined simulator such as information related to virtual space. For example, the information processing apparatus 100 may generate the prediction information based on area B in the virtual space that reflects the flow of people in area B and context information of the features of area B.

Also, the model information storage unit 125 of the information processing apparatus 100 may store a model applied to the agent in the simulation (hereinafter also referred to as "agent model"). For example, the information processing device 100 may generate a model to be applied to the agent in the simulation. For example, the information processing device 100 generates a model based on user behavior information. For example, the information processing apparatus 100 may generate an agent model based on user behavior information in the A area. For example, the information processing apparatus 100 generates an agent model corresponding to each attribute based on user behavior information. For example, the information processing apparatus 100 may generate various agent models such as a male agent model in her 20s and a female agent model in her 20s based on the behavior information of the user. For example, the information processing apparatus 100 may generate an agent model in consideration of various environmental information such as open data. For example, the information processing apparatus 100 may store the generated agent model in the model information storage unit 125 .

For example, the information processing apparatus 100 may generate simulation information (simulation log) regarding the action of the agent in the predetermined simulator using various information regarding the predetermined simulator such as information regarding the virtual space stored in the simulator information storage unit. good. For example, the information processing apparatus 100 may assign an agent model stored in the model information storage unit 125 to an agent, and generate simulation information (simulation log) by simulating push notification and its opening. The information processing device 100 may also generate prediction information based on the generated simulation information.

For example, the information processing apparatus 100 may generate a simulation log by simulating the behavior of a user (human) in area B, which is a prediction target area. For example, due to reasons such as insufficient number of push notifications to actual users in area B, prediction information may be generated based on simulations to predict the open rate of push notifications in area B. .

For example, the information processing apparatus 100 determines an agent model to apply to each agent. For example, the information processing apparatus 100 may determine an agent model to be applied to each agent based on information related to the population composition of the B area, etc., included in the environment information ED of the B area. For example, the information processing apparatus 100 may determine an agent model to be applied to each agent based on the demographic composition of area B regarding age and gender as attribute information. Attribute information used for population composition is not limited to age and sex, and various attribute information may be used.

In the following example, it is assumed that the population composition of area B is mostly men in their twenties. For example, the information processing apparatus 100 generates agents as many as the number of people (population, etc.) in the B area. In the example of FIG. 11, the information processing apparatus 100 assumes that the population in area B is 50,000 and that 50,000 agents are generated. For example, the information processing apparatus 100 assigns attributes to 50,000 agents so as to reflect the population composition in the A area. For example, if the proportion of men in their 20s in area B is 50%, the information processing apparatus 100 assigns the attribute of men in their 20s to 25,000 agents. Further, for example, when the percentage of women in their 20s in area B is 20%, the information processing apparatus 100 assigns the attribute of women in their 20s to 10,000 agents.

For example, the information processing apparatus 100 applies an agent model corresponding to a man in his twenties to an agent with the attribute of a man in his twenties. For example, the information processing apparatus 100 applies an agent model corresponding to a man in his twenties to 25,000 agents assigned the attribute of a man in his twenties. Further, for example, the information processing apparatus 100 applies an agent model corresponding to a woman in her twenties to 10,000 agents assigned the attribute of a woman in her twenties.

For example, the information processing apparatus 100 performs simulation using an agent to which an agent model is applied. For example, each agent applied to the virtual space corresponding to area B behaves according to its assigned agent model. For example, when an agent model for men in their twenties is applied, when a push notification is sent at a predetermined timing, depending on the context such as location and situation in area B at that time, whether to open the push notification or not. decide what to do. In addition, for example, an agent to which the agent model for women in their 20s is applied, when a push notification is sent at a predetermined timing, opens the push notification according to the context such as the location and situation in area B at that time. decide whether to

In this way, the information processing apparatus 100 generates a simulation log by causing each agent to act in the virtual space corresponding to the B area. For example, the information processing device 100 may store the generated simulation log in the storage unit 120 . For example, the information processing device 100 generates prediction information based on the generated simulation log. For example, the information processing apparatus 100 generates prediction information (prediction information PM11 in FIG. 11, etc.) with the B area as the prediction target area based on the simulation log corresponding to the B area.

[1-2. About model generation]
The information processing apparatus 100 may generate a model by various methods without being limited to the above. For example, the information processing apparatus 100 may generate (update) a model using information of the generated model. The information processing apparatus 100 may generate (update) the model by feeding back the information of the generated model. For example, the information processing apparatus 100 may generate (update) a model by using information of the generated model as a constraint. For example, the information processing apparatus 100 may generate (update) a model using a feature amount that satisfies a predetermined condition among the generated model information as a constraint condition.

For example, the information processing apparatus 100 generates a model including feature amounts excluding feature amounts whose absolute value of weight is equal to or less than a predetermined threshold value (for example, 0.01 or 0.001) among the information of the generated model. You can (update). For example, the information processing apparatus 100 removes the feature amount (feature amount #X) whose weight absolute value is 0.00001 and is equal to or less than a predetermined threshold value (0.001) from the generated model M1. may be generated (updated). For example, when the feature quantity #X corresponds to the combination of the elements “soccer field” and “cucumber”, the information processing apparatus 100 removes the feature quantity corresponding to the combination of the elements with a low relationship, thereby converting the model M1 into It can be generated (updated) more appropriately.

[2. Configuration of Information Processing Device]
Next, the configuration of the information processing apparatus 100 according to the embodiment will be described using FIG. FIG. 3 is a diagram illustrating a configuration example of an information processing apparatus according to the embodiment; As shown in FIG. 3, the information processing apparatus 100 has a communication section 110, a storage section 120, and a control section . The information processing apparatus 100 includes an input unit (for example, a keyboard, a mouse, etc.) that receives various operations from an administrator of the information processing apparatus 100, and a display unit (for example, a liquid crystal display, etc.) for displaying various information. may have.

(Communication unit 110)
The communication unit 110 is realized by, for example, a NIC (Network Interface Card) or the like. The communication unit 110 is connected to a network (for example, network N in FIG. 2) by wire or wirelessly, and transmits and receives information to and from the terminal device 10 .

(storage unit 120)
The storage unit 120 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. The storage unit 120 according to the embodiment, as shown in FIG. , a prediction information storage unit 126 and a relation information storage unit 127 .

Although not shown, the open data storage unit 121 stores various open data. For example, the open data storage unit 121 stores open data including traffic information such as road information such as construction and congestion, weather information such as weather and temperature, and economic information such as household budget and consumption. may

Although not shown, the storage unit 120 includes a content information storage unit that stores content information such as the content CT11. The content information storage unit stores tag information (for example, soccer, cucumber, etc.) indicated in each content in association with information identifying the content (content ID, etc.). The content information storage unit stores information about various contents such as the content itself such as the content CT11 and information indicating the category of the content such as the content CT11 (for example, sports, entertainment, etc.). Although not shown, the storage unit 120 includes a POI information storage unit that stores the facility name of each POI such as a store or a stadium, the location of the facility, the area, and the like. The POI information storage unit stores various types of information such as the category of each POI in association with information for identifying each POI (POI ID, etc.).

(User information storage unit 122)
The user information storage unit 122 according to the embodiment stores various information about users. For example, the user information storage unit 122 stores various types of information regarding user attributes and various types of information regarding user behavior. 4 is a diagram illustrating an example of a user information storage unit according to the embodiment; FIG. The user information storage unit 122 shown in FIG. 4 includes items such as "user ID", "age", "gender", "home", "place of work", "interest", and "location information".

"User ID" indicates identification information for identifying a user. "Age" indicates the age of the user identified by the user ID. Note that the “age” may be a specific age of the user identified by the user ID, such as 35 years old. "Gender" indicates the gender of the user identified by the user ID.

"Home" indicates location information of the home of the user identified by the user ID. In the example shown in FIG. 4, "home" is illustrated as an abstract code such as "LC11", but may be information indicating latitude and longitude. Also, for example, "home" may be a region name or an address.

Also, "work place" indicates the location information of the work place of the user identified by the user ID. In the example shown in FIG. 4, the "place of work" is illustrated as an abstract code such as "LC12", but may be information indicating latitude and longitude. Also, for example, the "place of work" may be an area name or an address.

"Interest" indicates the interest of the user identified by the user ID. That is, "interest" indicates an object in which the user identified by the user ID has a high interest. In the example shown in FIG. 4, one "interest" is shown for each user, but there may be more than one.

"Positional information" indicates user's positional information at a predetermined date and time (for example, when prediction information is generated). For example, "location information" indicates the latest location information of the user acquired before a predetermined date and time (for example, when the prediction information is generated). In the example of FIG. 4, "location information" is illustrated by abstract code "LC1-1" or the like, but specific latitude/longitude information or the like may be stored in "location information".

For example, in the example shown in FIG. 4, the age of the user identified by the user ID "U1" is "twenties" and the gender is "male". Also, for example, the user identified by the user ID "U1" indicates that the home is "LC11". Also, for example, the user identified by the user ID "U1" indicates that the place of work is "LC12". Also, for example, the user identified by the user ID “U1” indicates that he is interested in “economy”. For example, for user U1, the location information indicating that user U1 is located within partial area LC1-1 of area A is location information last acquired before a predetermined date and time, that is, the latest location information.

It should be noted that the user information storage unit 122 may store various types of information, not limited to the above, depending on the purpose. For example, the user information storage unit 122 may store other demographic attribute information and psychographic attribute information in addition to age and gender as user attribute information. For example, the user information storage unit 122 may store information such as name, family structure, and income.

(Action information storage unit 123)
The behavior information storage unit 123 according to the embodiment stores various types of information about user behavior. 5 is a diagram illustrating an example of a behavior information storage unit according to the embodiment; FIG. For example, the behavior information storage unit 123 stores user behavior information regarding content provided to the terminal device 10 of each user. The action information storage unit 123 shown in FIG. 5 includes items such as "user ID", "action ID", "type", "content", "date and time", and "position". Note that the behavior of the user is not limited to the behavior actively performed by the user, and may include various behaviors such as the behavior performed by the user, that is, the passive behavior of the user.

"User ID" indicates identification information for identifying a user. For example, a user identified by user ID "U1" corresponds to user U1 shown in the example of FIG. Also, the "action ID" indicates information for identifying the action of the user. "Type" indicates information about the type of the corresponding user's action. Also, "content" indicates the content targeted in the corresponding user's behavior. "Date and time" indicates the date and time when the corresponding user action was performed. The “date and time” is illustrated abstractly such as “dt11”, but a specific date and time such as “September 13, 2018, 22:31:52” may be stored.

"Position" indicates the user's position information corresponding to the user's behavior. For example, "position" indicates the user's position information when the user's behavior information was acquired. In the example of FIG. 5, "position" is illustrated by abstract code "LC1-1" or the like, but specific latitude/longitude information or the like may be stored in "position information". In addition, the "location information" may include POI (point of interest) names, areas, and the like, which indicate facility names such as shops and stadiums.

For example, in the example of FIG. 5, the user (user U1) identified by the user ID "U1" performs the action (action AC11) identified by the action ID "AC11" or the action (action) identified by the action ID "AC12". Indicates that action AC12) or the like has been performed. Specifically, the example of FIG. 5 shows that the content CT11 was notified (action AC11) to the user U1 located in the partial area LC1-1 on the date and time dt11. For example, action AC11 indicates that content CT11 was provided to the terminal device 10 used by user U1 on date and time dt11 by push notification. In the example of FIG. 5, the user U1 performed an operation (action AC12) to open the content CT11 in the partial area LC1-1 at date and time dt12 after the push notification of the content CT11 was made at date and time dt11. . Also, in the example of FIG. 5, the user U1 browsed the content CT11 (action AC13) during the time RTM1 (for example, 5 minutes, 10 minutes, etc.). For example, it indicates that user U1 browsed content CT11 (action AC13) in partial area LC1-1 over time RTM1 from date/time dt12 to date/time dt13.

Note that the behavior information storage unit 123 may store various types of information, not limited to the above, depending on the purpose. For example, the action information storage unit 123 may store position information of the user when each piece of action information was acquired in association with each piece of action information. Also, FIG. 5 shows the case where the action information is stored in the action information storage unit 123 for each user ID, but the action information is not limited to each user ID and may be stored, for example, in chronological order.

(Distribution information storage unit 124)
The distribution information storage unit 124 according to the embodiment stores various information regarding distribution. 6 is a diagram illustrating an example of a distribution information storage unit according to the embodiment; FIG. For example, the distribution information storage unit 124 stores distribution information of each tally target area. The distribution information storage unit 124 shown in FIG. 6 includes items such as "total target area" and "distribution information". The "distribution information" includes items such as "partial area", "push count", "open count", and "open rate (%)".

"Aggregation target area" indicates an area from which distribution information is collected. Also, "distribution information" indicates distribution information collected about the aggregation target area. A “partial area” indicates a partial area within the aggregation target area. "Number of pushes" indicates the number of push notifications sent to users located in the corresponding partial area. Also, the "number of times opened" indicates the number of times the user located in the corresponding partial area has opened the push notification. The "opening rate (%)" indicates the percentage of push notifications opened by users located in the corresponding partial area (=number of times of opening/number of times of pushing×100). For example, the "opening rate (%)" is a value between 0 and 100 (%).

For example, in the example shown in FIG. 6, it indicates that the distribution information has been collected for the aggregation target area "A area". The total target area “A area” corresponds to area AR11 in FIG. For example, it indicates that "10000 (times)" push notifications have been sent to the users located in the partial area LC1-1 within the aggregation target area "A area". Also, for example, it indicates that the user located in the partial area LC1-1 within the aggregation target area "A area" has opened the push notification "2000 (times)". Further, for example, the opening rate of users located in the partial area LC1-1 within the tally target area "A area" is "20 (%)" because 2,000 out of 10,000 notifications have been opened. indicates that

It should be noted that the distribution information storage unit 124 may store various types of information, not limited to the above, depending on the purpose.

(Model information storage unit 125)
The model information storage unit 125 according to the embodiment stores information about models. For example, the model information storage unit 125 stores model information used for generating prediction information. 7 is a diagram illustrating an example of a model information storage unit according to the embodiment; FIG. The model information storage unit 125 shown in FIG. 7 stores, as model information, items such as "feature #1 (cafe A x soccer)" to "feature #3 (cafe A x economy)" corresponding to the model M1 and the like. have

For example, in the example shown in FIG. 7, the model information about the model M1 is such that the weight of the feature value #1 corresponding to the combination of cafe A and soccer is "0.5", and the feature value #1 corresponding to the combination of cafe A and performing arts is "0.5". 2 is "-0.1", the weight of feature #3 corresponding to the combination of cafe A and economy is "0.2", and so on. For example, when the feature (feature amount) of the model is represented by an m-dimensional vector, the number of feature amounts (feature number) is m, and the weights of the feature amount ##1 to feature amount #m are stored. For example, the weight of each feature amount may be a real number or a negative value.

It should be noted that the model information storage unit 125 may store various types of model information according to purposes, not limited to the above. For example, the model information storage unit 125 may store various model information such as probability models.

(Prediction information storage unit 126)
The prediction information storage unit 126 according to the embodiment stores various information related to prediction. For example, the prediction information storage unit 126 stores various information related to prediction and various information generated by prediction. 8 is a diagram illustrating an example of a prediction information storage unit according to the embodiment; FIG. The prediction information storage unit 126 shown in FIG. 8 includes items such as “prediction target area” and “prediction information”. The "prediction information" includes items such as "partial area", "push count", "open count", and "open rate (%)".

“Prediction target area” indicates an area for which prediction information is generated. Also, "prediction information" indicates prediction information generated for the prediction target area. "Partial area" indicates a partial area within the prediction target area. The “number of pushes” indicates the number of push notifications predicted to be sent to users located in the corresponding partial area in the generation of prediction information. In addition, the "number of times of opening" indicates the number of times it is predicted that the push notification will be opened by the user located in the corresponding partial area in the generation of the prediction information. In addition, the "opening rate (%)" indicates the rate (=number of times of opening/number of times of pushing×100) that users located in the corresponding partial area are predicted to open the push notification in the generation of prediction information. For example, the "opening rate (%)" is a value between 0 and 100 (%).

For example, in the example shown in FIG. 8, it indicates that the prediction information has been generated for the prediction target area “B area”. The prediction target area “B area” corresponds to area AR21 in FIG. For example, it indicates that it is predicted that “2000 (times)” push notifications will be sent to users located in the partial area LC2-1 within the prediction target area “B area”. Further, for example, it is predicted that the user located in the partial area LC2-1 within the prediction target area “B area” will open the push notification “40 (times)”. Further, for example, the opening rate of users located in the partial area LC2-1 in the prediction target area "B area" is predicted to be 40 times out of 2,000 notifications. )”.

Note that the prediction information storage unit 126 may store various types of information, not limited to the above, depending on the purpose.

(Relationship information storage unit 127)
The relational information storage unit 127 according to the embodiment stores various types of information regarding elements related to predetermined actions of the user. For example, the relationship information storage unit 127 stores various types of information related to the feature amount related to the user's predetermined behavior among the feature amounts of the model that predicts the user's predetermined behavior. 9 is a diagram illustrating an example of a relationship information storage unit according to the embodiment; FIG. The relational information storage unit 127 shown in FIG. 9 has items such as “target behavior”, “order”, “relational information ID”, “weight”, “element #1”, “element #2”, and “element #3”. included. The relationship information storage unit 127 includes not only “element #1” to “element #3” but also “element #4”, “element #5”, and other items corresponding to the number of elements corresponding to the feature amount. be

"Target behavior" indicates a predetermined behavior of a target user. "Rank" indicates the rank (ranking) of the degree of relationship between each piece of related information and the target action. In the example of FIG. 9, "ranking" indicates a case where a higher rank (ranking) is assigned as the absolute value of the weight of each piece of relational information increases. "Relationship information ID" indicates information for identifying the relationship information. "Weight" indicates the weight of the feature quantity corresponding to the relationship information. "Element #1" to "Element #3" and the like indicate the elements of the feature quantity corresponding to the relationship information.

The example shown in FIG. 9 indicates that the target action is the user's action on the content. For example, it indicates that the target action is a user's action such as opening a content. Further, the relationship information (relationship information ST1) identified by the relationship information ID “ST1” corresponds to the feature quantity with the weight of “0.5”, the element #1 is “Cafe A”, and the element #2 is Indicates that it is "soccer". That is, the relationship information ST1 corresponds to the feature amount #1 of the model M1.

Note that the relationship information storage unit 127 may store various types of information, not limited to the above, depending on the purpose.

(control unit 130)
Returning to the description of FIG. 3, the control unit 130 is a controller, and is stored in a storage device inside the information processing apparatus 100 by, for example, a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). Various programs (corresponding to an example of a determination program or an information processing program) are executed by using the RAM as a work area. Also, the control unit 130 is a controller, and is implemented by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 3, the control unit 130 includes an acquisition unit 131, a determination unit 132, a generation unit 133, and a provision unit 134, and implements or executes the information processing functions and actions described below. . Note that the internal configuration of the control unit 130 is not limited to the configuration shown in FIG. 3, and may be another configuration as long as it performs information processing described later.

(Acquisition unit 131)
Acquisition unit 131 acquires various types of information. The acquisition unit 131 acquires various types of information from an external device such as the terminal device 10 . Acquisition unit 131 acquires various types of information from storage unit 120 . Acquisition unit 131 acquires various types of information from an external information processing device. The acquisition unit 131 acquires various types of information from the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, the distribution information storage unit 124, the model information storage unit 125, the prediction information storage unit 126, and the relationship information storage unit 127. get. The acquisition unit 131 acquires behavior information of a predetermined user. The acquisition unit 131 acquires information about open data. The acquisition unit 131 acquires behavior information such as position information from the terminal device 10 .

Acquisition unit 131 acquires context information about distribution of content and content information about content. The acquisition unit 131 acquires a plurality of combination information indicating each combination of a plurality of first elements of context information and a plurality of second elements of content information. The acquisition unit 131 acquires a model that is used for predicting a predetermined behavior of the user with respect to content, using each combination of the plurality of first elements and the plurality of second elements as each feature amount.

Acquisition unit 131 acquires a model generated based on correct information, which is action information related to a predetermined action of a user, context information, and content information for content distributed to a user located in a predetermined area. . The acquisition unit 131 acquires a model generated based on correct information, which is action information related to a user's predetermined action, context information, and content information for content distributed to the user. The acquisition unit 131 acquires a model that includes each feature amount corresponding to each element and that is used for predicting a predetermined behavior of the user with respect to the content based on the context information of the context.

Acquisition unit 131 acquires other action information related to predetermined actions of other users with respect to content distributed to other users located in other areas. The acquisition unit 131 acquires other action information related to a predetermined action of another user with respect to content distributed to another user located in another area different from the predetermined area.

The acquisition unit 131 acquires action information such as operation information indicating whether or not a user performs a predetermined operation on content distributed to the terminal device 10 used by the user located in a predetermined area. The acquisition unit 131 acquires action information such as operation information indicating whether or not the user has performed a predetermined operation on the content notified by push in the terminal device used by the user located in the predetermined area. The acquisition unit 131 acquires action information such as operation information indicating whether or not the user has opened the content, which is notified by push notification on the terminal device used by the user located in the predetermined area.

The acquisition unit 131 acquires action information such as operation information associated with location information of the user when the user performs a predetermined operation on the content notified by push within a predetermined area. The acquisition unit 131 acquires action information such as operation information associated with time information when the user performs a predetermined operation on the content notified by push within a predetermined area. The acquisition unit 131 acquires action information such as operation information associated with context information when the user performs a predetermined operation on content that has been pushed within a predetermined area. The acquisition unit 131 acquires action information such as operation information associated with environment information related to a predetermined area when the user performs a predetermined operation on the content notified by push notification in the predetermined area.

In the example of FIG. 1, the acquisition unit 131 acquires information indicating that predetermined content has been pushed to the terminal device 10 used by a certain user and has been opened by the user. Acquisition unit 131 acquires information indicating that predetermined content has been pushed to the terminal device 10 used by user U1 located in area A and has been opened by user U1. Acquisition unit 131 acquires information indicating that predetermined content has been pushed to terminal device 10 used by user U1 located in partial area LC1-1 of area A and has been opened by user U1.

The acquisition unit 131 acquires information indicating that the notified content (content CT11) has been viewed by the user for the time RTM1 after opening. The acquisition unit 131 may acquire information indicating that the content notified in step S1-1 was displayed on the terminal device 10 of the user for the time RTM1 after the content was opened.

The acquisition unit 131 acquires information indicating that predetermined content has been pushed to the terminal device 10 used by a certain user and has not been opened by the user. The acquisition unit 131 acquires information indicating that predetermined content was pushed to the terminal device 10 used by the user U2 located in the A area and was not opened by the user U2. The acquisition unit 131 acquires information indicating that predetermined content was pushed to the terminal device 10 used by the user U2 located in the partial area LC1-2 of the A area and was not opened by the user U2.

The acquisition unit 131 acquires open data of the prediction target area. In the example of FIG. 11, the acquisition unit 131 acquires the environment information ED of the B area. The acquisition unit 131 may acquire various information such as environment information of the B area from the open data storage unit 121 (see FIG. 3).

In addition, the acquisition unit 131 acquires user information located in the prediction target area. In the example of FIG. 11, the acquisition unit 131 acquires the user information UD of the B area. The acquisition unit 131 acquires user information such as that shown in the user information storage unit 122 (see FIG. 4) and user behavior information such as that shown in the behavior information storage unit 123 (see FIG. 5).

(Determination unit 132)
The determination unit 132 determines various types of information. The determination unit 132 determines various information. The determination unit 132 determines various information based on the various information stored in the storage unit 120 . The determination unit 132 determines various information based on various information stored in the storage unit 120 . Based on the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, the distribution information storage unit 124, the model information storage unit 125, the prediction information storage unit 126, the relationship information storage unit 127, etc. , determine various information. Based on the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, the distribution information storage unit 124, the model information storage unit 125, the prediction information storage unit 126, the relationship information storage unit 127, etc. , determine various information. The determination unit 132 determines various information based on the various information acquired by the acquisition unit 131 . The determination unit 132 determines various information based on the various information acquired by the acquisition unit 131 . The determination unit 132 determines various information based on the various information generated by the generation unit 133 . The determination unit 132 determines various information based on the various information generated by the generation unit 133 . The determination unit 132 determines a model to be used for generating prediction information. In the example of FIG. 11, the determining unit 132 determines the model M11 as the model to be used for generating prediction information.

In the example of FIG. 1, the determination unit 132 determines the following based on each combination of the element group EL1 corresponding to the context regarding content distribution and the element group EL2 corresponding to the details of the content such as the content category among the context information. , determine the features of the model. For example, the determining unit 132 determines each combination of the first element of the element group EL1 and the second element of the element group EL2 as the feature amount of the model. For example, the determination unit 132 determines each combination of the first element of the element group EL1, such as “Cafe A”, and the second element of the element group EL2, such as “soccer” and “entertainment”, as the feature quantity of the model. to decide.

The determining unit 132 determines feature quantity #1 corresponding to the combination of cafe A and soccer, feature quantity #2 corresponding to the combination of cafe A and entertainment, feature quantity #3 corresponding to the combination of cafe A and economy, and the like. The feature amount of the model M1 is determined.

The determination unit 132 provides content to the user based on the generated relationship information. First, the determination unit 132 determines a user (target user) to whom content is to be provided. The determination unit 132 determines the user U3 located at the store SP1, which is the cafe A in the B area, as the target user. The determining unit 132 determines the user U3 who is the second user located in the B area, which is an area other than the A area, as the target user. The determination unit 132 determines content to be provided to the user U3 who is the target user, based on the generated relationship information. The determination unit 132 determines to provide soccer-related content (soccer-related content CT31) to the user U3 located at the location “Cafe A”.

The determination unit 132 receives user information, environmental information about the user's location, and the like, and the score output by the model M11 based on the input is a predetermined threshold value (for example, "0.5", etc.) or more. In this case, it may be determined that the user opens the content for which the push notification is sent. The determining unit 132 inputs various information including position information and behavior information of the user U3 to the model M11, and determines that the user U3 does not open the push notification when the model M11 outputs a score less than a predetermined threshold. You may The determining unit 132 inputs various information including position information and behavior information of the user U4 to the model M11, and determines that the user U4 opens the push notification when the model M11 outputs a score equal to or higher than a predetermined threshold. may

The determination unit 132 determines an agent model to be applied to each agent. The determination unit 132 may determine an agent model to be applied to each agent based on information regarding the population composition of area B, etc., included in the environment information ED of area B. FIG. The determining unit 132 may determine an agent model to be applied to each agent based on the demographic composition of area B regarding age and gender as attribute information.

(Generating unit 133)
The generation unit 133 generates various types of information. The generator 133 estimates various types of information. The generator 133 selects various types of information. The generation unit 133 extracts various information. The generator 133 calculates various information. The generation unit 133 estimates various information based on various information stored in the storage unit 120 . The generation unit 133 selects various information based on various information stored in the storage unit 120 . The generation unit 133 extracts various information based on various information stored in the storage unit 120 . The generation unit 133 calculates various information based on various information stored in the storage unit 120 . Based on the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, the distribution information storage unit 124, the model information storage unit 125, the prediction information storage unit 126, the relationship information storage unit 127, and the like, the generation unit 133 , estimate various information. Based on the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, the distribution information storage unit 124, the model information storage unit 125, the prediction information storage unit 126, the relationship information storage unit 127, and the like, the generation unit 133 , select various information. Based on the open data storage unit 121, the user information storage unit 122, the behavior information storage unit 123, the distribution information storage unit 124, the model information storage unit 125, the prediction information storage unit 126, the relationship information storage unit 127, and the like, the generation unit 133 , extract various information. The generation unit 133 estimates various information based on the various information acquired by the acquisition unit 131 . The generation unit 133 selects various information based on the various information acquired by the acquisition unit 131 . The generation unit 133 extracts various information based on the various information acquired by the acquisition unit 131 . The generation unit 133 uses the information acquired by the acquisition unit 131 to estimate the user's context.

The generation unit 133 generates various types of information stored in the storage unit 120 . The generation unit 133 generates distribution information stored in the distribution information storage unit 124 . The generation unit 133 generates a model stored in the model information storage unit 125 . The generation unit 133 generates prediction information stored in the prediction information storage unit 126 .

Based on the context information and the content information acquired by the acquisition unit 131, the generation unit 133 generates relationship information indicating a combination of the first element of the context information related to content distribution and the second element of the content information. Generate. The generation unit 133 generates relationship information indicating a combination of the first element and the second element regarding the position of the delivery destination of the content. The generation unit 133 generates relationship information indicating a combination of the first element and the second element regarding the delivery time of the content. The generating unit 133 generates relational information indicating a combination of the first element regarding the position and the second element of the information regarding the details of the content.

The generation unit 133 generates relationship information indicating a combination of the first element and the second element by selecting combination information related to content distribution from among a plurality of pieces of combination information. The generation unit 133 generates relationship information indicating a combination of the first element and the second element by selecting a feature amount related to distribution of the content among the feature amounts of the model. The generation unit 133 generates relationship information by selecting feature amounts whose corresponding weights satisfy predetermined conditions from among the feature amounts of the model. The generation unit 133 generates relational information by selecting feature amounts for which the absolute value of the corresponding weight is equal to or greater than a predetermined threshold among the feature amounts of the model.

The generation unit 133 generates relationship information indicating a combination of the first element and the second element corresponding to another area different from the predetermined area. The generating unit 133 generates related element information based on the model. The generating unit 133 generates relationship information indicating a combination of the first element and the second element by selecting the feature amount related to content distribution among the feature amounts of the model based on other behavior information. Generate. If the score based on the other behavior information is equal to or greater than a predetermined threshold, the generation unit 133 selects the feature amount related to content distribution from among the feature amounts of the model, thereby generating the first element and the second element. Generate relationship information indicating a combination with .

In the example of FIG. 1, the generation unit 133 generates model information of the model M1 as shown in the model information MD. For example, the generation unit 133 generates the model M1 as shown in the model information storage unit 125 (see FIG. 7). The generation unit 133 generates the model M1 by learning the weight of each feature amount such as the feature amount #1 to the feature amount #3 of the model M1.

The generation unit 133 generates relationship information indicating a combination of the first element and the second element based on the generated model M1. Based on the generated model M1, the generation unit 133 generates relationship information indicating the correspondence between the combination of the first element and the second element and the degree of relationship between the user's opening action for the content. Based on the generated model M1, the generation unit 133 generates relationship information indicating the degree of relationship between the first element corresponding to the context related to distribution of the content, the second element corresponding to the content, and the browsing behavior of the user for the content. do.

For example, the generation unit 133 generates relationship information indicating the degree of relationship between the content and the predetermined behavior of the user based on the weight of each feature amount of the model M1. The generation unit 133 generates relationship information by selecting, from among the feature amounts of the model M1, a feature amount whose corresponding weight satisfies a predetermined condition. The generation unit 133 generates relationship information by selecting a feature amount for which the absolute value of the corresponding weight is greater than a predetermined threshold among the feature amounts of the model M1. In the example of FIG. 1, the generation unit 133 generates relational information by selecting, from among the feature amounts of the model M1, the feature amount whose corresponding weight absolute value is greater than the threshold "0.3". Note that the threshold is not limited to "0.3" and may be set as appropriate. Further, for example, the generating unit 133 generates relational information in which relational information corresponding to each feature quantity is ranked such that a feature quantity having a large absolute value of weight among the feature quantities of the model M1 is ranked high. do.

The generation unit 133 selects, from among the relational information stored in the relational information storage unit 127, the relational information having elements such as “Cafe A” and “Cafe” relating to the place. For example, the generation unit 133 selects a predetermined number (for example, 5 or 10) of relational information items having the highest order among the relational information items having location-related elements such as “cafe A” and “cafe”. In the example of FIG. 1, the generation unit 133 selects the relational information ST1, the relational information ST4, etc., including the element "Cafe A" in descending order of rank.

Based on the action information such as the operation information acquired by the acquisition unit 131, the generation unit 133, when the content is distributed to the terminal device 10 used by the user located in another area different from the predetermined area, Prediction information for predicting whether or not the user will perform a predetermined operation on the content is generated. Based on the action information such as the operation information acquired by the acquisition unit 131, the generation unit 133 pushes the content to a terminal device used by the user located in another area different from the predetermined area. Prediction information for predicting whether or not the user will perform a predetermined operation on the content is generated. Based on the action information such as the operation information acquired by the acquisition unit 131, the generation unit 133 generates a push notification of content to a terminal device used by a user located in another area. Prediction information for predicting the presence or absence of an opening operation is generated.

The generation unit 133 generates prediction information for predicting a position at which the user performs a predetermined operation on content to be notified by push notification in another area, based on action information such as operation information. The generation unit 133 generates prediction information for predicting the time when the user performs a predetermined operation on the content to be notified by push notification in another area based on the action information such as the operation information.

The generation unit 133 predicts whether or not the user performs a predetermined operation on the content to be notified by push notification in the other area based on the action information such as the operation information and the context information of the user located in the other area. Generate forecast information. The generation unit 133 generates prediction information for predicting whether or not the user has performed a predetermined operation on content to be notified by push notification in another area, based on behavior information such as operation information and environment information about the other area. do.

The generating unit 133 generates, as prediction information, distribution information indicating a distribution of users who perform predetermined operations on content to be notified by push notifications in other areas, based on action information such as operation information. The generation unit 133 generates, as prediction information, a distribution map indicating the distribution in other areas of users who perform predetermined operations with respect to content to be notified by push notifications in other areas, based on action information such as operation information. .

The generation unit 133 generates distribution information based on action information such as operation information related to push notifications. In the example of FIG. 11, the generation unit 133 generates the distribution information DM11 with the A area as the aggregation target area. The generation unit 133 generates a model used for generating prediction information based on action information such as operation information related to push notification. The generating unit 133 generates a model for predicting whether to open the push notification based on action information such as operation information in the A area.

The generation unit 133 generates a model for predicting whether to open the push notification based on various information stored in the open data storage unit 121, the user information storage unit 122, the action information storage unit 123, and the like. The generation unit 133 generates a model based on the environment information of the A area stored in the open data storage unit 121 . The generation unit 133 generates a model based on the user information of the user located in the A area stored in the behavior information storage unit 123 . The generation unit 133 generates a model based on the behavior information of the user located in the A area stored in the user information storage unit 122 . The generating unit 133 may generate a model by taking into account information about content to be pushed to the user.

The generation unit 133 may generate a model for predicting whether or not to open the push notification, using information when the content notified by the push notification is opened as correct information (positive example). The generation unit 133 opens the push notification using various context information such as the user information of the user, the user's location information, the user's behavior information, and the user's environment information when the content to be notified of the push is opened as correct information. You may generate a model that predicts whether or not When the user U1 opens the content of which the push notification is sent in the partial area LC1-1, the generation unit 133 generates various information such as the user information of the user U1, the location information of the user U1, the behavior information of the user U1, and the environment information of the user U1. may be used as correct information to generate a model that predicts whether or not to open the push notification.

The generation unit 133 may also generate a model for predicting whether or not to open the push notification, using information when the content notified by the push notification has not been opened as incorrect answer information (negative example). The generation unit 133 generates push notifications using various context information such as user information, user location information, user behavior information, and user environment information when the content to be notified by push notification is not opened as incorrect answer information. may generate a model that predicts whether to open If the user U2 does not open the content to which the push notification is sent in the partial area LC1-2, the generation unit 133 generates the user information of the user U2, the location information of the user U2, the behavior information of the user U2, the environment information of the user U2, and the like. may be used as incorrect information to generate a model that predicts whether to open the push notification.

In the example of FIG. 11, the generation unit 133 generates model information of the model M11 as shown in model information MD. The generation unit 133 generates the model M11 as shown in the model information storage unit 125 (see FIG. 7). The generation unit 133 receives as input context information including user information and environmental information around the user's position, and generates a model M11 that outputs a score based on the input. The generation unit 133 receives the context information and generates a model M11 that outputs a score based on the input. The generation unit 133 generates prediction information based on the generated model M11. The generation unit 133 generates prediction information about the B area based on the model M11 and various types of information about the B area, which is the prediction target area.

The generation unit 133 may generate prediction information for users located in the B area during a predetermined time period. The generation unit 133 may generate prediction information corresponding to area B in the morning, targeting users located in area B in the morning time period (6:00 to 10:00). The generation unit 133 may generate prediction information corresponding to the B area at night, targeting users located in the B area during the night time period (19:00 to 23:00).

In the example of FIG. 11, the generation unit 133 generates prediction information PM11 with the B area as the prediction target area. The generation unit 133 generates prediction information that predicts the push notification opening rate for area B, where the push notification record is insufficient.

(Providing unit 134)
The providing unit 134 provides various information. The providing unit 134 provides various types of information to an external information processing device. The providing unit 134 provides various information to the terminal device 10 . The providing unit 134 provides various types of information to an external information processing device such as the terminal device 10 . The providing unit 134 transmits various types of information to the terminal device 10 . The providing unit 134 distributes various types of information to the terminal device 10 . The providing unit 134 provides various information based on the various information acquired by the acquiring unit 131 . The providing unit 134 provides various information based on the various information determined by the determining unit 132 . The providing unit 134 provides various information based on the various information generated by the generating unit 133 . The providing unit 134 may provide various services based on the prediction information generated by the generating unit 133 .

In the example of FIG. 1, the providing unit 134 provides content to the user based on the generated relationship information. The providing unit 134 provides the soccer-related content CT31 to the user U3. The providing unit 134 provides the soccer-related content CT31 to the terminal device 10-3 used by the user U3. The providing unit 134 transmits the soccer-related content CT31 to the terminal device 10-3 used by the user U3. The providing unit 134 distributes the soccer-related content CT31 to the terminal device 10-3 used by the user U3.

The providing unit 134 may perform push notification based on the generated prediction information. In the example of FIG. 11, the providing unit 134 may send a push notification to a user located in a heavily hatched portion (lower right in area B) in the prediction information PM11. The providing unit 134 may provide the generated prediction information to an external information processing device. The providing unit 134 may provide prediction information to an external device that provides a content providing service.

[3. Information processing flow]
Next, the procedure of information processing by the information processing system 1 according to the embodiment will be described with reference to FIG. 10 . FIG. 10 is a flowchart illustrating an example of information processing according to the embodiment;

As shown in FIG. 10, the information processing apparatus 100 acquires context information regarding distribution of content (step S101). Then, the information processing apparatus 100 acquires content information regarding the content (step S102). Then, based on the context information and the content information, the information processing apparatus 100 generates relationship information indicating a combination of the first element of the context information and the second element of the content information (step S103).

[4. effect〕
As described above, the information processing apparatus 100 according to the embodiment has the acquisition unit 131 and the generation unit 133 . Acquisition unit 131 acquires context information about distribution of content and content information about content. Based on the context information and the content information acquired by the acquisition unit 131, the generation unit 133 generates relationship information indicating a combination of the first element of the context information related to content distribution and the second element of the content information. Generate.

As described above, the information processing apparatus 100 according to the embodiment generates the first element of the context information related to the delivery of the content and the second element of the content information based on the context information related to the delivery of the content and the content information related to the content. By generating relationship information indicating a combination of , it is possible to appropriately generate information indicating the relationship between context and content regarding distribution of content.

Further, in the information processing apparatus 100 according to the embodiment, the generation unit 133 generates relationship information indicating a combination of the first element and the second element regarding the position of the distribution destination of the content.

As described above, the information processing apparatus 100 according to the embodiment generates relationship information indicating a combination of the first element and the second element regarding the position of the distribution destination of the content. can appropriately generate information indicating the relationship between

In addition, in the information processing apparatus 100 according to the embodiment, the generation unit 133 generates relationship information indicating a combination of the first element and the second element regarding the delivery time of the content.

As described above, the information processing apparatus 100 according to the embodiment generates relational information indicating a combination of the first element and the second element regarding the time of content distribution, thereby providing context regarding content distribution and content. Information indicating relationships can be generated appropriately.

In addition, in the information processing apparatus 100 according to the embodiment, the generation unit 133 generates relationship information indicating a combination of a first element related to position and a second element of information related to content details.

As described above, the information processing apparatus 100 according to the embodiment generates context related to content delivery by generating relationship information indicating a combination of a first element related to position and a second element of information related to content details. It is possible to appropriately generate information indicating the relationship between content and content.

Also, in the information processing apparatus 100 according to the embodiment, the acquisition unit 131 acquires a plurality of combination information indicating each combination of the plurality of first elements of the context information and the plurality of second elements of the content information. The generation unit 133 generates relationship information indicating a combination of the first element and the second element by selecting combination information related to content distribution from among a plurality of pieces of combination information.

As described above, the information processing apparatus 100 according to the embodiment selects combination information related to distribution of content from among a plurality of pieces of combination information, thereby obtaining relationship information indicating a combination of the first element and the second element. By generating it, it is possible to appropriately generate information indicating the relationship between the context regarding the distribution of the content and the content.

In addition, in the information processing apparatus 100 according to the embodiment, the acquisition unit 131 uses each combination of a plurality of first elements and a plurality of second elements as each feature amount, and obtains a model to be used for predicting a predetermined behavior of the user with respect to content. get. The generation unit 133 generates relationship information indicating a combination of the first element and the second element by selecting a feature amount related to distribution of the content among the feature amounts of the model.

As described above, the information processing apparatus 100 according to the embodiment uses each combination of a plurality of first elements and a plurality of second elements as each feature amount, and each feature amount of a model used for predicting a predetermined behavior of a user with respect to content. Information indicating the relationship between the context regarding the delivery of the content and the content by selecting the feature amount related to the delivery of the content, and generating relationship information indicating the combination of the first element and the second element. can be generated properly.

Further, in the information processing apparatus 100 according to the embodiment, the generation unit 133 generates relationship information by selecting a feature amount whose corresponding weight satisfies a predetermined condition from among the feature amounts of the model.

As described above, the information processing apparatus 100 according to the embodiment selects, from among the feature amounts of the model, feature amounts whose corresponding weights satisfy predetermined conditions, thereby generating relationship information, thereby distributing content. Information that indicates the relationship between context and content can be generated appropriately.

Further, in the information processing apparatus 100 according to the embodiment, the generation unit 133 generates relationship information by selecting, from each feature amount of the model, a feature amount whose corresponding weight absolute value is equal to or greater than a predetermined threshold value. do.

As described above, the information processing apparatus 100 according to the embodiment selects feature amounts whose corresponding weight absolute values are equal to or greater than a predetermined threshold among the feature amounts of the model, thereby generating relationship information. , it is possible to appropriately generate information indicating the relationship between the content and the context regarding the delivery of the content.

Further, in the information processing apparatus 100 according to the embodiment, the acquisition unit 131 generates content distributed to the user based on the correct answer information, which is action information related to a predetermined action of the user, the context information, and the content information. Get the model that was created.

As described above, the information processing apparatus 100 according to the embodiment can generate a model generated based on the correct information, which is action information related to a predetermined action of the user, the context information, and the content information for the content distributed to the user. By generating the related element information based on, it is possible to appropriately generate information indicating the relationship between the context regarding the distribution of the content and the content.

Further, in the information processing apparatus 100 according to the embodiment, the acquisition unit 131 obtains correct information, context information, and content information, which are action information related to a predetermined action of the user, for content distributed to a user located in a predetermined area. to get a model generated based on The generation unit 133 generates relationship information indicating a combination of the first element and the second element corresponding to another area different from the predetermined area.

In this way, the information processing apparatus 100 according to the embodiment can combine correct information, which is behavior information related to a predetermined behavior of a user, context information, and content information with respect to content distributed to a user located in a predetermined area. Based on the model generated based on the model, by generating relationship information indicating the combination of the first element and the second element corresponding to another area different from the predetermined area, context related to content distribution and content can appropriately generate information indicating the relationship between For example, the information processing apparatus 100 uses information of a predetermined area (first area) with sufficient information to generate prediction information of a user located in another area (second area) with insufficient information. allows the desired prediction even for areas with insufficient information.

Further, in the information processing apparatus 100 according to the embodiment, the acquisition unit 131 is a model that includes each feature amount corresponding to each element, and is used to predict a predetermined behavior of the user with respect to content based on context information of the context. Get the model to use. The generating unit 133 generates related element information based on the model.

As described above, the information processing apparatus 100 according to the embodiment is a model that includes each feature amount corresponding to each element, and is a model that is used to predict a predetermined behavior of a user with respect to content based on the context information of the context. By generating relational element information based on this, it is possible to appropriately generate information indicating the relationship between the context regarding content distribution and the content.

[5. Hardware configuration]
The information processing apparatus 100 according to the embodiments described above is implemented by a computer 1000 configured as shown in FIG. 12, for example. FIG. 12 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing apparatus. The computer 1000 includes a CPU 1100, a RAM 1200, a ROM (Read Only Memory) 1300, a HDD (Hard Disk Drive) 1400, a communication interface (I/F) 1500, an input/output interface (I/F) 1600, and a media interface (I/F). ) 1700.

The CPU 1100 operates based on programs stored in the ROM 1300 or HDD 1400 and controls each section. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, a program depending on the hardware of the computer 1000, and the like.

The HDD 1400 stores programs executed by the CPU 1100, data used by the programs, and the like. Communication interface 1500 receives data from other devices via network N, sends the data to CPU 1100, and transmits data generated by CPU 1100 to other devices via network N. FIG.

The CPU 1100 controls output devices such as displays and printers, and input devices such as keyboards and mice, through an input/output interface 1600 . CPU 1100 acquires data from an input device via input/output interface 1600 . CPU 1100 also outputs the generated data to an output device via input/output interface 1600 .

Media interface 1700 reads programs or data stored in recording medium 1800 and provides them to CPU 1100 via RAM 1200 . CPU 1100 loads such a program from recording medium 1800 onto RAM 1200 via 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) or a PD (Phase change rewritable disc), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. etc.

For example, when the computer 1000 functions as the information processing apparatus 100 according to the embodiment, the CPU 1100 of the computer 1000 implements the functions of the control unit 130 by executing programs loaded on the RAM 1200 . The CPU 1100 of the computer 1000 reads these programs from the recording medium 1800 and executes them, but as another example, these programs may be acquired via the network N from another device.

As described above, some of the embodiments of the present application have been described in detail based on the drawings, but these are merely examples, and various modifications, It is possible to carry out the invention in other forms with modifications.

[6. others〕
Further, among the processes described in the above embodiments, all or part of the processes described as being automatically performed can be manually performed, or the processes described as being performed manually can be performed manually. All or part of this can also be done automatically by known methods. In addition, information including processing procedures, specific names, various data and parameters shown in the above documents and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each drawing is not limited to the illustrated information.

Also, each component of each device illustrated is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution and integration of each device is not limited to the one shown in the figure, and all or part of them can be functionally or physically distributed and integrated in arbitrary units according to various loads and usage conditions. Can be integrated and configured.

Further, each process described in each embodiment described above can be appropriately combined within a range in which the contents of the process are not inconsistent.

Also, the above-mentioned "section, module, unit" can be read as "means" or "circuit". For example, the acquisition unit can be read as acquisition means or an acquisition circuit.

1 information processing system 100 information processing device 121 open data storage unit 122 user information storage unit 123 behavior information storage unit 124 distribution information storage unit 125 model information storage unit 126 prediction information storage unit 127 relationship information storage unit 130 control unit 131 acquisition unit 132 Determining unit 133 Generating unit 134 Providing unit 10 Terminal device 50 Information providing device N Network

Claims (12)

an acquisition unit that acquires context information about distribution of content and content information about the content;
relational information indicating a combination of a first element of the context information and a second element of the content information related to distribution of the content, based on the context information and the content information obtained by the obtaining unit; a generation unit that generates relationship information used to provide target content corresponding to the second element combined with the first element to a target user corresponding to the first element;
with
The acquisition unit
obtaining a plurality of combination information indicating each combination of the plurality of first elements of the context information and the plurality of second elements of the content information;
The generating unit
An information processing apparatus, wherein the relationship information indicating the combination of the first element and the second element is generated based on the plurality of pieces of combination information. The generating unit
2. The information processing apparatus according to claim 1, wherein said relationship information indicating a combination of said first element and said second element relating to the position of said distribution destination of said content is generated. The generating unit
2. The information processing apparatus according to claim 1, wherein said relationship information indicating a combination of said first element and said second element relating to time of distribution of said content is generated. The generating unit
4. The information processing according to any one of claims 1 to 3, wherein relationship information indicating a combination of said first element relating to position and said second element of information relating to content of said content is generated. Device. The generating unit
The relationship information indicating the combination of the first element and the second element is generated by selecting combination information related to distribution of the content from among the plurality of combination information. 5. The information processing device according to any one of 1 to 4. The acquisition unit
Acquiring a model to be used for predicting a predetermined behavior of a user with respect to the content, using each combination of the plurality of first elements and the plurality of second elements as each feature amount,
The generating unit
The relationship information indicating the combination of the first element and the second element is generated by selecting a feature amount related to distribution of the content from among the feature amounts of the model. The information processing device according to claim 5 . The generating unit
7. The information processing apparatus according to claim 6, wherein the relationship information is generated by selecting the feature amount whose corresponding weight satisfies a predetermined condition from among the feature amounts of the model. The generating unit
8. The relationship information is generated by selecting, from among the feature amounts of the model, the feature amount for which the absolute value of the corresponding weight is equal to or greater than a predetermined threshold. The information processing device according to . The acquisition unit
Acquiring the model generated based on correct answer information, which is action information related to a predetermined action of the user with respect to the content distributed to the user, the context information, and the content information. The information processing apparatus according to any one of claims 6 to 8. The acquisition unit
The model generated based on the correct answer information, which is the action information related to the user's predetermined action, the context information, and the content information with respect to the content delivered to the user located in the predetermined area. and get
The generating unit
10. The information processing apparatus according to claim 9, wherein said relationship information indicating a combination of said first element and said second element corresponding to another area different from said predetermined area is generated. A computer-executed information processing method comprising:
an obtaining step of obtaining context information about distribution of content and content information about the content;
relational information indicating a combination of a first element of the context information and a second element of the content information related to distribution of the content, based on the context information and the content information obtained by the obtaining step; a generating step of generating relationship information used to provide a target user corresponding to the first element with target content corresponding to the second element combined with the first element;
including
The obtaining step includes
obtaining a plurality of combination information indicating each combination of the plurality of first elements of the context information and the plurality of second elements of the content information;
The generating step includes
An information processing method, comprising generating the relationship information indicating a combination of the first element and the second element based on the plurality of pieces of combination information. an acquisition procedure for acquiring context information relating to delivery of content and content information relating to said content;
relationship information indicating a combination of a first element of the context information related to distribution of the content and a second element of the content information, based on the context information and the content information acquired by the acquisition procedure; a generation procedure for generating relationship information used to provide target content corresponding to the second element combined with the first element to a target user corresponding to the first element;
on the computer, and
The acquisition procedure includes:
obtaining a plurality of combination information indicating each combination of the plurality of first elements of the context information and the plurality of second elements of the content information;
The generating procedure includes:
An information processing program, wherein the relationship information indicating the combination of the first element and the second element is generated based on the plurality of pieces of combination information.

Images