JP7001508B2 - Information processing equipment, information processing methods, and programs. - Google Patents

Description

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

Conventionally, a determination unit that determines the order of a plurality of contents including articles, each of which is textual information about a predetermined target, and a determination unit that determines the order based on the user information of the user are displayed one by one in the order determined by the determination unit. The user's terminal displays the plurality of contents in which the next-ranked content of the content is displayed in accordance with the reception of the user's evaluation for each content whose evaluation can be specified. A distribution device including a distribution unit for distribution to the device is disclosed (see, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2016-189206

However, in the above-mentioned conventional technique, the content to be provided to the user is appropriately determined and provided by using the location information.

The present invention has been made in consideration of such circumstances, and one of the objects of the present invention is to provide an information processing device, an information processing method, and a program capable of providing a more suitable content to a user. do.

One aspect of the present invention includes an acquisition unit that acquires the position information of the first user, an estimation unit that estimates the behavior attribute of the first user based on the position information acquired by the acquisition unit, and the estimation unit. It is an information processing apparatus including a providing unit that provides the content corresponding to the behavior attribute of the first user estimated by the estimation unit to the first user.

According to one aspect of the present invention, more suitable content can be provided to the user.

It is a figure which shows an example of the usage environment in which an information providing apparatus is used. It is a figure which shows an example of the functional structure of an information processing apparatus 10. It is a figure which shows an example of the flow of the process executed by the information processing apparatus 10. It is a figure which shows an example of the contents of the position history information 32. It is a figure which shows the distribution of the location information, and the estimated home and work. It is a figure which shows an example of the content of the behavior analysis information 36. It is a figure which shows an example of the content of the action attribute information 38. It is a flowchart which shows an example of the flow of the process executed by the determination unit 22. It is a figure which shows an example of the content of the extended information 44 in which a plurality of categories are associated with the behavior attribute of a user. It is a flowchart (the 1) which shows an example of the flow of the process executed by the information processing apparatus 10 of the 2nd Embodiment. 2 is a flowchart (No. 2) showing an example of a flow of processing executed by the information processing apparatus 10 of the second embodiment. It is a figure for demonstrating change of a user's behavior attribute or a category. This is another example of the content of the behavior attribute information 38A.

Hereinafter, embodiments of the information processing apparatus, information processing method, and program of the present invention will be described with reference to the drawings.

The information processing device is realized by one or more processors. The information processing device acquires the position information of the first user, estimates the behavior attribute of the first user based on the acquired position information, and provides the content corresponding to the estimated behavior attribute of the first user. Provide to the first user. User behavior attributes are user behavior attributes derived from location information, not static attributes such as gender, occupation, and age, but user interests, interests, lifestyles, life stages, and consumption trends. Of these, it indicates one or more elements.

For example, if the information processing device estimates that the user visits the real estate agent at a predetermined frequency based on the location information, the user presumes that he / she is interested in the real estate, and the content according to the user's interest. To the user. The content may be any content related to the user's behavior attribute. For example, if the user is interested in real estate, the content is information that the user likes, such as an advertisement for real estate or news related to real estate.

Further, the information processing apparatus further acquires the behavior history of the plurality of users in the network, extracts the user having the same behavior history in the network as the first user included in the plurality of users as the second user, and first. The content corresponding to the behavior attribute of the user is provided to the second user. As a result, the information processing apparatus presumes that the second user having a similar behavior history has the same interests and interests as the first user, and the content corresponding to the estimated interests and interests is the first. Provide to 2 users. This makes it possible to provide the user with more suitable content.

<First Embodiment>
[Constitution]
FIG. 1 is a diagram showing an example of a usage environment in which an information providing device is used. The usage environment of the information providing device includes, for example, terminal devices 2-1 to 2-N (“N” is an arbitrary natural number) and server devices 4-1 to 4-n (“n” is an arbitrary natural number). , The advertisement distribution device 6 and the information processing device 10. The information processing device 10 may include the advertisement distribution device 6. Further, hereinafter, when the terminal devices 2-1 to 2-N are not distinguished, they are simply referred to as "terminal device 2", and when the server devices 4-1 to 4-n are not distinguished, they are simply referred to as "server device 4". Refer to.

These devices and the like communicate with each other via the network NW. The network NW includes, for example, a WAN (Wide Area Network), a LAN (Local Area Network), the Internet, a dedicated line, a wireless base station, a provider, and the like.

[Terminal device]
The terminal device 2 is, for example, a mobile phone such as a smartphone or a portable device such as a tablet terminal. The terminal device 2 includes a positioning device (not shown), a control unit, a communication unit, a display unit, an operation unit, and the like. The positioning device includes a GPS (Global Positioning System) receiver. The positioning device performs positioning based on the radio wave received from the satellite by the GPS receiver, and identifies the position of the terminal device 2 (that is, the position of the user). Further, the terminal device 2 may estimate the position of the terminal device 2 from the position of the radio base station to which the communication unit is connected.

[Server device]
The plurality of server devices 4 are provided for each service provided to the user, for example. The server device 4 provides a predetermined service to the terminal device 2 that has accessed the server device 4. For example, the server device 4 acquires a user ID and location information from the accessing terminal device 2, and based on the acquired location information, provides information on recommended stores existing in the vicinity or information on the surrounding weather. Or provide. For example, when the user operates the operation unit of the terminal device 2 to access the server device 4, information such as a user ID and location information registered in advance is transmitted to the server device 4. The server device 4 associates the acquired user ID, position information, the time when these information are acquired, etc. with each other and stores them in the storage unit as position history information, and transmits the stored position history information to the information processing device 10. ..

The above-mentioned user ID is, for example, a user ID commonly used in a plurality of server devices 4. The user (for example, the first user) who uses the terminal device 2 has the same service provided by, for example, the server device 4-1 and the service provided by the server device 4-n among the plurality of server devices 4. It can be used with a user ID.

Further, the server device 4 acquires a query from the terminal device 2, searches for information related to the acquired query, and provides the search result to the terminal device 2. The query information acquired by the server device 4 from the terminal device 2 is provided to the information processing device 10. The server device 4 may request another device to search for information related to the acquired query, and may acquire the result of the search by the other device.

[Advertising distribution device]
The advertisement distribution device 6 provides an advertisement to the server device 4, for example, based on a request from the server device 4. The advertisement provided is, for example, an advertisement (content) according to the behavior attribute of the user determined by the information processing apparatus 10. In the present embodiment, it is assumed that the advertisement distribution device 6 is provided, but instead of (or in addition to) the advertisement distribution device 6 may be provided with a content distribution device that distributes content other than advertisements. In this case, the content distribution device provides the user with content according to the user's behavior attribute determined by the information processing device 10.

[Information processing device]
FIG. 2 is a diagram showing an example of the functional configuration of the information processing apparatus 10. The information processing apparatus 10 includes, for example, a communication unit 12, a position information acquisition unit 14, a first estimation unit 16, a second estimation unit 18, a processing unit 20, a determination unit 22, a provision unit 24, and storage. A unit 30 is provided. In the position information acquisition unit 14, the first estimation unit 16, the second estimation unit 18, the processing unit 20, the determination unit 22, and the provision unit 24, for example, a hardware processor such as a CPU (Central Processing Unit) is stored in a storage device. It is realized by executing the stored program. Further, these functional units may be realized by hardware such as LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), GPU (Graphics Processing Unit), etc. It may be realized by the collaboration of software and hardware. Further, the above program may be stored in a storage device in advance, or is stored in a removable storage medium such as a DVD or a CD-ROM, and the storage medium is attached to the drive device of the information processing device 10. It may be installed in the storage device.

The storage unit 30 is realized by, for example, a ROM (Read Only Memory), a flash memory, an SD card, a RAM (Random Access Memory), an HDD (Hard Disc Drive), a register, or the like. Further, a part or all of the storage unit 30 may be NAS, an external storage server, or the like. The storage unit 30, for example, stores position history information 32, map information 34, behavior analysis information 36, behavior attribute information 38, query information 40, and an extended model 42. The details of the information stored in the storage unit 30 will be described later.

The communication unit 12 includes a communication interface such as a network interface card.

The position information acquisition unit 14 stores the position history information 32 acquired from the server device 4 via the communication unit 12 in the storage unit 30.

The first estimation unit 16 estimates the user's home or workplace based on the position information acquired by the position information acquisition unit 14. The second estimation unit 18 estimates the user's behavior attribute based on the estimation result of the first estimation unit 16 and the position information acquired by the position information acquisition unit 14. The combination of the first estimation unit 16 and the second estimation unit 18 is an example of the "estimation unit".

The processing unit 20 extracts a second user who is presumed to have the same or similar behavior attribute as the first user's behavior attribute based on the estimated user behavior attribute and the behavior history in the user's network. Generate an extended model 42 to do so.

The determination unit 22 determines and extracts as the second user a user whose behavior history in the network is similar to that of the first user included in the plurality of users. The determination unit 22 is an example of the “extraction unit”.

The providing unit 24 provides the content according to the behavior attribute of the first user to the first user or the second user.

[flowchart]
FIG. 3 is a diagram showing an example of a flow of processing executed by the information processing apparatus 10. First, the first estimation unit 16 selects a user to be processed from the position history information 32 (S10). Next, the first estimation unit 16 acquires the position information of the selected user (S12). FIG. 4 is a diagram showing an example of the contents of the position history information 32. As shown in FIG. 4, the position history information 32 is information in which a user ID, a history of position information of a user having the user ID, and a time associated with the position information are associated with each other.

(Processing to estimate home and work)
Next, the first estimation unit 16 estimates the user's home or workplace based on the acquired position information (S14). FIG. 5 is a diagram showing the distribution of location information and the estimated home and work. The first estimation unit 16 specifies, for example, the position of the user between 2:00 and 5:00 at midnight, and if the same position information exists in a predetermined number or more in the time zone in a week, the position is estimated as home. Further, for example, the first estimation unit 16 specifies the position of the user between 10:00 and 12:00 in the daytime, and if the same position information exists in a predetermined number or more in the time zone in a week, the position is estimated as a workplace. do.

Further, when estimating the home or the workplace, the type of the building corresponding to the location information included in the map information 34 may be added. For example, the type of building corresponding to the user's position between 2:00 and 5:00 at midnight is an office building, and the type of building corresponding to the user's position between 10:00 and 12:00 in the daytime is a residential building such as an apartment. For example, it may be estimated that the position of the user between 2 o'clock and 5 o'clock in the middle of the night is the office and the position of the user between 10 o'clock and 12 o'clock in the daytime is the home. In this way, the home and the workplace are estimated based on the location information.

Further, the first estimation unit 16 may perform clustering on the position information by using a predetermined clustering method, and estimate the position satisfying the predetermined standard as home or work. The predetermined clustering method is a Gaussian Mixture Model (GMM), a K-means method, DBSCAN, or the like. The method for specifying the home or the workplace is not limited to the above method, and various known methods may be used or may be given in advance.

(Process to analyze behavior)
Next, the second estimation unit 18 analyzes the user's behavior based on the estimated home, work location, and other location information (S16). For example, the second estimation unit 18 refers to the position history information 32 and the map information 34, identifies the time when the user is at home or work, or is present at another place, estimates the user's behavior, and acts. The analysis information 36 is generated.

FIG. 6 is a diagram showing an example of the contents of the behavior analysis information 36. The behavior analysis information 36 is information indicating the user's behavior including the place where the user stopped by, the place where the user stayed, and the like. For example, the second estimation unit 18 identifies the facility or store where the user has stopped by based on the location information and the information such as the facility or store associated with the location information registered in the map information 34. By the above processing, as shown in the upper figure of FIG. 6, the user "ID001" spent the day at work from Monday to Friday, spent the night at home from Monday to Friday, and spent one day on Saturday. It is estimated that he was staying at home and spent Sunday at A theme park. In addition, by the above processing, as shown in the lower figure of FIG. 6, the user "ID002" spent the day at work from Monday to Friday, spent the night at home from Monday to Friday, and 1 on Saturday. It is highly probable that he was staying at home for the day and staying at a real estate agent on Sunday afternoon.

(Process to determine behavior attributes)
Next, the second estimation unit 18 determines the behavior attribute of the user based on the behavior analysis information 36 as shown in FIG. 6 and the behavior attribute information 38 (S18). For example, the second estimation unit 18 determines whether or not the content of the behavior analysis information 36 satisfies the condition of the behavior attribute information 38, and determines the behavior attribute of the user based on the determination result. Further, for example, the second estimation unit 18 estimates the behavior attribute of the user by using the position of one or both of the user's home and work, or excluding the position of one or both of the home and work. Or, the behavior attribute of the user is estimated based on the information of the facility corresponding to the location information.

FIG. 7 is a diagram showing an example of the contents of the behavior attribute information 38. The behavior attribute information 38 is information in which identification information, behavior attributes, and conditions for classifying users are associated with each other. Behavioral attributes include, for example, working in a workplace with long working hours (identification information "01"), being interested in real estate (identification information "02"), and being interested in travel (identification information "01"). 03 "), being interested in sports (identification information" 04 "), being interested in theme parks (identification information" 05 "), not being interested in exercise (identification information" 06 "), and the like.

For example, the condition classified into the behavior attribute of the identification information "01" is, for example, staying at the workplace for XX hours or more per week, and the condition classified into the behavior attribute of the identification information "02" is the month. The condition that the person has visited the real estate agent more than 〇〇 times and is classified into the behavioral attribute of the identification information “06” is that the time spent at home is XX hours or more. The above conditions and the conditions shown in FIG. 7 are examples, and the behavior attributes classified when the conditions in which the above conditions are combined are set, or when the above conditions and other conditions are satisfied, the behavior attributes are classified. May be set. For example, the staying time at home is XX hours, the staying time at work is XX hours, and the staying time (location information) of both the home and the workplace may be used for the determination of classification. ..

Machine learning or deep learning technology may be used for the process of determining the behavior attribute of the user. For example, when the position information for a predetermined period of time for the user who wants to classify the behavior attribute is input to the trained model, it is determined whether or not the input position information corresponds to the predetermined behavior attribute. This trained model is a model trained using position information that is known to be classified into a predetermined behavioral attribute as teacher data. The teacher data may include location information (negative example) that is known not to be classified into a given behavioral attribute.

Next, the processing unit 20 determines whether or not a predetermined number of user behavior attributes have been specified (S20). If the behavior attributes of a predetermined number of users have not been specified, the process returns to S10. It should be noted that, instead of whether or not a predetermined number of user behavior attributes have been specified, each of the predetermined behavior attributes (for example, the behavior attributes of the identification information "01" to "06") has been processed by the predetermined number. It may be determined whether or not it is associated with the user.

(Process to get query)
When a predetermined number of user behavior attributes are specified, the processing unit 20 acquires a query of the user whose behavior attribute is specified from the query information 40 (S22). The query information 40 is information acquired from the server device 4, and is information including a query given to the server device 4 by the user in the past.

(Process to extract characteristic queries)
Next, the processing unit 20 extracts a characteristic query from the query acquired in S22 (S24). For example, the processing unit 20 uses TF-IDF (Term Frequency-Inverse Document Frequency) or the like to narrow down important queries from the queries acquired in S22 and extract characteristic queries. It should be noted that the processing of S24 may be omitted and the query may not be narrowed down.

(Process to generate extended model)
Next, the processing unit 20 generates the extended model 42 by machine learning (S26). The extended model 42 is a model used for estimating a user's behavior attribute from a feature amount of a behavior history (for example, a query) used in the processing of the flowchart of FIG. 8 to be described later. For example, the processing unit 20 performs machine learning using the feature amount of the user to be learned as correct answer data, and determines whether or not the feature amount of the user to be expanded and the feature amount of the user to be learned are similar. Generate the extended model 42. The extended model 42 is generated for each behavior attribute of the user. For example, the first extension model 42 is generated based on the information of the learning target user having the first behavior attribute, and the second extension is based on the information of the learning target user having the second behavior attribute. Model 42 is generated.

The feature amount handled by the processing unit 20 is, for example, in a vector in which each of a plurality of queries is associated with an element (for example, a tens of thousands of dimensional vectors having tens of thousands of elements), the query corresponding to each element is a user. Is a vector whose element value is 1 if is a query entered in the past, and zero if it is not. This vector is sometimes referred to as a sparse vector because the queries entered by the user are not exhaustive. Hereinafter, this is referred to as a user vector. The user vector is an example of "user feature quantity".

The processing unit 20 generates some trained model by machine learning so that the score value becomes high when the user vector of the user to be learned is input. The trained model is generated, for example, as a component such as SVM (Support Vector Machine), logistics regression analysis, and neural network. This trained model is, for example, a model trained using the user vector of the user whose predetermined behavior attribute is determined by the process of determining the behavior attribute as teacher data. The teacher data may include a user vector (negative example) that is known not to be classified into a given behavioral attribute.

When the user vector of the user who has entered a query similar to the query entered by the user to be learned is input to the extended model 42, the extended model 42 is similar to the case where the user vector of the user to be learned is input. Output a high score value. Since the extended target user who obtained this high score value is a user who has a user vector close to the learning target user in the vector space, the behavior attribute of the expansion target user is similar to the behavior attribute of the learning target user. It is inferred to do.

In addition, the determination unit 22 instructs the advertisement distribution device 6 to provide the user with the content corresponding to the behavior attribute to the user whose behavior attribute is determined in S18. This ends the processing of one routine in this flowchart.

By the above-mentioned processing, the information processing apparatus 10 can generate the extended model 42 for each behavior attribute of the user. Further, the information processing apparatus 10 can determine the behavior attribute of the user based on the position information and provide appropriate content to the user based on the determined behavior attribute.

[Extension of users by extension model]
FIG. 8 is a flowchart showing an example of the flow of processing executed by the determination unit 22. By this process, the behavior attribute of the user whose behavior attribute has not been determined is determined. First, the determination unit 22 selects a user (extended target) whose behavior attribute has not been determined (S30), and acquires a query for the selected user (S32). The user whose action attribute has not been determined is a user who has not been processed in the process of FIG. For example, a user whose location information has not been acquired or has not been sufficiently acquired.

Next, the determination unit 22 generates a user vector based on the acquired query (S34). Next, the determination unit 22 inputs the generated user vector to the expansion model 42, and determines the behavior attribute of the user to be expanded based on the result output by the expansion model 42 (S36). For example, the determination unit 22 includes a first user vector (first feature amount) showing the characteristics of the behavior history in the network of the user (first user) whose behavior attribute is determined based on the position information, and a plurality of users. A second user vector (second feature amount) showing the characteristics of the behavior history in the network of the included users is extracted, and a user having a second user vector similar to the first user vector is extracted. Then, the behavior attribute of the user having the second user vector is classified into the behavior attribute of the user having the first user vector.

When the behavior attribute is determined using the expansion model 42, for example, in S36, the determination unit 22 inputs the user vector of the user to be expanded into the plurality of expansion models 42 generated for each behavior attribute of the user. The behavior attribute associated with the extended model 42 that outputs the highest score is determined as the behavior attribute of the user.

Next, the providing unit 24 instructs the advertisement distribution device 6 to provide the content corresponding to the behavior attribute determined in S36 to the extended target user whose behavior attribute is determined (S38). The advertisement distribution device 6 distributes the content according to the determined behavior attribute to the user based on the instruction of the providing unit 24. For example, the advertisement distribution device 6 distributes content to the user based on the content information. The content information is information in which the behavior attribute and the content to be delivered are associated with each other. This ends the processing of one routine in this flowchart.

By the process described above, it is possible to determine the behavior attribute of the user who has not acquired the location information or sufficiently acquire it, and provide the user with the content corresponding to the determined behavior attribute. As a result, more appropriate content can be provided to the user more easily.

In the above example, the extended model 42 is derived by using the user vector based on the query (or keyword), but instead of (or in addition to) the query, the purchase history and the server device 4 are used. A user vector generated based on the browsing history for the provided information may be used as a feature of the user.

Further, the determination unit 22 acquires the result related to the advertisement distribution after delivering the content corresponding to the determined action attribute (or category) to the user, and if the acquired result is less than or equal to a predetermined degree, the determined action attribute. (Or category) may be changed. The acquired result is less than or equal to a predetermined degree, for example, a CTR (Click Through Rate) indicating the ratio of the number of times the content provided to the user is clicked to the number of times the content is displayed on the display unit of the user's terminal device 2. The CVR (Conversion Rate), which indicates the ratio of the number of times the product (or service) included in the content is purchased, etc., out of the number of times displayed on the display unit of the user's terminal device 2, is not more than a predetermined degree. When the acquired result is less than or equal to a predetermined degree, the determination unit 22 determines, for example, the behavior attribute associated with the extended model 42 that outputs the second highest score among the scores output by the extended model 42 described above. The user's behavior attributes may be changed.

[Example of extended information]
In S26 of the flowchart of FIG. 3, the processing unit 20 substitutes (or additionally) for the generation of the extended model 42, and the extended information 44 in which the user's behavior attribute, the query extracted in S24, and the category are associated with each other. May be generated. The extended information 44 is information used for estimating a user's behavior attribute or determining a category from a behavior history (for example, a query).

The processing unit 20 may associate a plurality of categories with the behavior attributes of the user. For example, in users with the same behavior attribute, a category is set for each query classified according to a predetermined criterion. FIG. 9 is a diagram showing an example of the contents of the extended information 44 in which a plurality of categories are associated with the user's behavior attribute. In the extended information 44 shown in FIG. 9, different categories are associated with each of the queries of the first group to the third group for the behavior attribute of one user. For example, in the behavior attribute of the identification information "01", the query of the first group includes a query related to work such as "working hours", "not resting", and "overtime", and the query of the second group includes "query". Queries related to health and body such as "health" and "good food" are included, and queries in the third group include queries related to how to enjoy holidays such as "travel", "drive", and "beautiful scenery ...". It is included. Then, different categories are associated with the first group to the third group.

In this case, in the process of S26 of FIG. 3 described above, the processing unit 20 determines the query (or keyword) included in each of the first group to the third group from the query acquired in S22, and the first group to the first group. The query included in each of the three groups is set in the extended information 44.

[Application example of extended information]
For example, in the processing of the flowchart of FIG. 8, the determination unit 22 may determine the behavior attribute or category of the user to be expanded by using the expansion information 44 instead of the expansion model 42. For example, instead of the processes of S34 and S36, the following processes are performed. In the extended information 44, the determination unit 22 is associated with a category to which a frequently input query (or a frequently input query, a frequently used keyword, a frequently input keyword, etc.) is associated with the extended information 44. Is determined as the category corresponding to the user. In this case, the advertisement distribution device 50 provides the user with the content according to the determined category.

Further, in the processing of the flowchart of FIG. 8, the determination unit 22 may determine the behavior attribute or category of the user to be expanded by using the expansion information 44 in addition to the expansion model 42. (1) For example, after the action attribute is determined by the extension model 42, the determination unit 22 associates the frequently entered query by the extension target user with the determined action attribute in the extension information 44. Determine if it is. (2) When associated, the determination unit 22 determines that the behavior attribute of the user is the behavior attribute determined by using the extended model 42. (3) When not associated, for example, the determination unit 22 generates a user vector for a query whose input frequency is a predetermined degree or more among the queries input by the user, and performs the above-mentioned processing. To determine the user's behavior attributes.

The extended information 44 may also be used in the process (S18) for determining the user's behavior attribute in the flowchart of FIG. In this case, for example, the determination unit 22 executes the same process as (1) to (3) above. Further, the determination unit 22 may determine the category (further subdivided behavior attribute) based on the extended information 44 and the frequently input query or the like by the user after determining the behavior attribute of the user. ..

As described above, by further using the extended information 44 in addition to the extended model 42, the behavior attribute of the user can be determined more accurately.

According to the first embodiment described above, the information processing apparatus 10 estimates the behavior attribute of the first user based on the position information of the first user, and sets it as the estimated behavior attribute of the first user. By providing the corresponding content to the first user, it is possible to provide the first user with more suitable content. Further, the information processing apparatus 10 further acquires the behavior history of the plurality of users in the network, extracts the user having the same behavior history in the network as the first user included in the plurality of users, and extracts the user as the second user. By providing the content corresponding to the behavior attribute of the first user to the second user, it is possible to provide the second user with more suitable content.

<Second Embodiment>
Hereinafter, the second embodiment will be described. In the second embodiment, when the POI (Point of Interest) information (POI information) of the facility or the like is updated in the map information 34, the action attribute (or category) determined for the user according to the update. Is changed. Hereinafter, the differences from the first embodiment will be described.

The second estimation unit 18 changes the behavior attribute of a specific user based on the updated map information 34. The providing unit 24 provides the content according to the behavior attribute of the first user to one or both of the first user and the second user, and the facility information (POI information) corresponding to the location information is changed. If so, one or both of the contents provided to the first user and the second user are changed.

FIG. 10 is a flowchart (No. 1) showing an example of the flow of processing executed by the information processing apparatus 10 of the second embodiment. First, the second estimation unit 18 determines whether or not a predetermined period has elapsed (S40). When the predetermined period has elapsed, the second estimation unit 18 acquires the updated POI information in the map information 34 (S42). As a result, the second estimation unit 18 was registered as a convenience store for the predetermined location information before the update, but after the update, the facility was changed to a real estate agent, and the facility became a different facility. You can recognize that it has changed. The second estimation unit 18 acquires the updated map information 34, compares the difference between the acquired updated map information 34 and the map information 34 before the update, and acquires the updated POI information. You may.

Next, the information processing apparatus 10 reflects the updated POI information acquired in S42 and executes the processes of S10 to S26 in the flowchart of FIG. 3 (S44). By the above processing, the extended model 42 reflecting the updated POI information is generated. This ends the processing of one routine in this flowchart.

By the above processing, even when the POI information is updated, an extended model capable of determining an appropriate behavior attribute for the user to be extended is generated.

FIG. 11 is a flowchart (No. 2) showing an example of the flow of processing executed by the information processing apparatus 10 of the second embodiment. First, the determination unit 22 determines whether or not the processing of the flowchart of FIG. 10 is executed and the extended model is updated (S50). When the extension model is updated, the determination unit 22 selects a user (expansion target) whose behavior attribute has not been determined (S52), and acquires a query for the selected user (S54). Next, the determination unit 22 generates a user vector based on the acquired query (S56).

Next, the determination unit 22 inputs the user vector generated in S56 into the updated expansion model 42, and determines the behavior attribute of the user to be expanded based on the result output by the updated expansion model 42. (S58). Next, the providing unit 24 instructs the advertisement distribution device 6 to distribute the content corresponding to the behavior attribute determined in S58 to the extended target user whose behavior attribute is determined (S60). This ends the processing of one routine in this flowchart.

As described above, even when the POI information is updated, by using the updated expansion model 42, an appropriate behavior attribute is determined for the user to be expanded, and the content corresponding to the determined behavior attribute is provided. Can be provided. As a result, it is possible to provide the user with more suitable content.

FIG. 12 is a diagram for explaining the change of the behavior attribute of the user. For example, it is assumed that the user is a user who is interested in real estate based on the map information 34 before the update, and the behavior attribute identification information "02" is specified (see FIG. 7).

In the above situation, when the POI information is updated and the processing unit 20 recognizes that the real estate agent visited by the user has been changed to a sports facility, the behavior analysis information 36 and the behavior attribute information 38 are referred to. Then, the behavior attribute of the user is changed to "interested in sports", and the extended model 42 is generated based on the changed behavior attribute and the user vector. Then, the determination unit 22 modifies the user's behavior attribute to the identification information "04" by using the extended model 42 and the user vector.

As described above, even when the POI information of the map information 34 is updated, the behavior attribute set for the user is changed according to the update of the POI information, so that the appropriate behavior attribute is reset for the user. be able to. Further, the updated expansion model 42 is also used for the user to be expanded. As a result, it is possible to provide the user with more suitable content.

In the above example, the extended model 42 is used, but instead of this, the behavior analysis information 36 and the extended information 44 may be updated by using the updated POI information. In this case, when the behavior analysis information 36 and the extended information 44 are updated, the determination unit 22 determines the behavior attribute and category of the user to be expanded by using the updated behavior analysis information 36 and the extended information 44.

According to the second embodiment described above, the information processing apparatus 10 provides one or both of the contents provided to the first user and the second user when the information of the facility corresponding to the location information is changed. By changing the map information 34, it is possible to provide the user with more suitable content even when the map information 34 is updated.

<Modification example>
In the modified example, the user's life stage, the user's consumption tendency, and the like are added to the user's behavior attribute. In addition to the behavioral attributes described with reference to FIG. 7 described above, the behavioral attribute information 38A may include, for example, an behavioral attribute indicating a user's life stage and an behavioral attribute indicating a user's consumption tendency. FIG. 13 is another example of the content of the behavior attribute information 38A. Behavioral attributes that indicate a user's life stage include, for example, having small children and having children attending a cram school (having children in elementary school or junior high school). Further, the behavioral attribute indicating the consumption tendency of the user is, for example, frequent visits to a toy store or frequent visits to a store offering a predetermined product (or service).

For example, the advertisement distribution device 6 provides a user with a small child with an advertisement for a product (for example, a picture book or a school bag) necessary for raising a small child or going on to school. Further, the advertisement distribution device 6 may provide the user with real estate information according to the life stage based on the behavior attribute indicating the life stage. For example, a user who is planning to have a child in the near future may be provided with an advertisement for a larger residence (for example, a residence where three family members can live). Further, the advertisement distribution device 6 provides the user with an advertisement for a new product toy, for example, to a user who frequently goes to a toy store.

As described above, more effective content can be provided to the user based on the life stage and the consumption tendency of the user.

According to each of the above-described embodiments, the information processing apparatus 10 first obtains the position information of the first user based on the position information acquisition unit 14 and the position information acquired by the position information acquisition unit 14. The estimation unit (first estimation unit 16 and second estimation unit 18) that estimates the behavior attribute of the user and the content corresponding to the behavior attribute of the first user estimated by the estimation unit are provided to the first user. By providing the providing unit 24, more suitable content can be provided to the first user.

Further, according to each of the above-described embodiments, the information processing apparatus 10 further includes a determination unit 22 for extracting as a second user a user whose behavior history in the network is similar to that of the first user included in the plurality of users. By providing the content corresponding to the behavior attribute of the first user to the second user, it is possible to provide the second user with more suitable content.

Although the embodiments for carrying out the present invention have been described above using the embodiments, the present invention is not limited to these embodiments, and various modifications and substitutions are made without departing from the gist of the present invention. Can be added.

2-1 to 2-N ... Terminal device, 4-1 to 4-n ... Server device, 6 ... Advertisement distribution device, 10 ... Information processing device, 12 ... Communication unit, 14 ... Location information acquisition unit, 16 ... 1st Estimating unit, 18 ... 2nd estimation unit, 20 ... processing unit, 22 ... decision unit, 24 ... providing unit, 30 ... storage unit, 32 ... position history information, 34 ... map information, 36 ... behavior analysis information, 38 ... behavior Attribute information, 40 ... query information, 42 ... extended model, 44 ... extended information

Claims (16)

An acquisition unit that acquires the location information of a plurality of first users,
An estimation unit that estimates each behavior attribute of the first user based on the position information of each of the first users acquired by the acquisition unit, and an estimation unit.
A decision unit that determines the behavior attribute associated with the first user having a behavior history similar to the behavior history in the network of the second user who cannot determine the behavior attribute based on the position information.
The content corresponding to the behavior attribute of the first user estimated by the estimation unit is provided to the first user.
A providing unit that provides the content according to the behavior attribute determined by the determining unit to the second user, and a providing unit.
Information processing device equipped with. The determination unit includes an action history in a network of a second user whose position information has not been acquired or a second user who cannot determine an action attribute based on the acquired position information, and a behavior history of the first user. Based on the behavior history, it is determined which behavior attribute the second user's behavior attribute is classified into among the behavior attributes of the first user.
The information processing apparatus according to claim 1. The behavioral attribute indicates one or more elements of the user's interest, interest, lifestyle, life stage, and consumption tendency.
The information processing apparatus according to claim 1 or 2. The estimation unit identifies one or both of the home and the workplace of the first user based on the location information, and estimates the behavior attribute of the first user by using the identified result. ,
The information processing apparatus according to any one of claims 1 to 3. The estimation unit estimates the behavior attribute of the first user by excluding the location information corresponding to one or both of the home and the workplace from the location information.
The information processing apparatus according to claim 4. The estimation unit estimates the behavioral attribute of the first user based on one or both of the time spent at the home and the time spent at the workplace based on the location information.
The information processing apparatus according to claim 4 or 5. The estimation unit estimates the behavior attribute of the first user based on the information of the facility corresponding to the location information.
The information processing apparatus according to any one of claims 1 to 6. The providing unit provides content according to the behavior attribute of the first user to one or both of the first user and the second user, and the information of the facility corresponding to the location information is changed. If so, the content provided to one or both of the first user and the second user is changed.
The information processing apparatus according to claim 1 or 2. The first feature amount showing the characteristics of the behavior history in the network of the first user and the second feature amount showing the characteristics of the behavior history in the network of the users included in the plurality of users are extracted and used as the first feature amount. It is provided with an extraction unit for extracting a user having a similar second feature amount as a second user.
The information processing apparatus according to claim 1 or 2. In the trained model in which the feature amount indicating the feature in the behavior history of the first user's network is learned, the third feature amount showing the feature in the behavior history of the user's network included in the plurality of users is input to the trained model. A extraction unit for determining whether or not a user included in the plurality of users corresponds to a second user based on the result output by the trained model is provided .
The information processing apparatus according to claim 1 or 2. The first user is a user who uses a service provided by a plurality of server devices by using common identification information.
The behavior history in the network is collected across a plurality of services.
The information processing apparatus according to claim 1, 2, 9, or 10. The behavioral attributes are that you spend a lot of time at work, are interested in real estate, are interested in travel, are interested in sports, are interested in theme parks, or are interested in exercise. One of the things that isn't there,
The information processing apparatus according to any one of claims 1 to 11. The decision-making part
A feature amount indicating a feature in the behavior history of the second user's network is input to each of the plurality of trained models prepared for each behavior attribute of the first user, and the result output by the trained model is obtained. Based on this, the first user having an action history similar to the action history of the second user is determined, and the action attribute associated with the first user is determined.
The trained model is a model in which a feature amount indicating a feature in the behavior history of the first user's network is trained.
The information processing apparatus according to any one of claims 1 to 8. The providing part
The content corresponding to the category corresponding to the behavior attribute of the first user estimated by the estimation unit and the behavior history in the network of the first user is provided to the first user.
The content corresponding to the category corresponding to the behavior attribute of the second user and the behavior history of the second user determined by the determination unit is provided to the second user.
The information processing apparatus according to any one of claims 1 to 13. The computer
Acquire the location information of multiple first users and
Based on the acquired position information of each of the first users , each behavior attribute of the first user is estimated.
The behavior attribute associated with the first user having a behavior history similar to the behavior history in the network of the second user who cannot determine the behavior attribute based on the position information is determined.
The content corresponding to the estimated behavior attribute of the first user is provided to the first user.
The content corresponding to the determined behavior attribute is provided to the second user .
Information processing method. On the computer
Acquire the location information of multiple first users,
Based on the acquired position information of each of the first users , each behavior attribute of the first user is estimated.
The behavior attribute associated with the first user having a behavior history similar to the behavior history in the network of the second user who cannot determine the behavior attribute based on the position information is determined.
The first user is made to provide the content corresponding to the estimated behavior attribute of the first user.
The second user is provided with the content corresponding to the determined behavior attribute .
program.

Images