JP2022087409A - Data addition system and data addition method - Google Patents

Abstract

An object of the present invention is to add data that can be used for marketing to potential users who are similar to a target user and whose individuals cannot be identified. [Solution] Data including a record of a potential user who is more than a certain degree similar to a target user including specified conditions and a record of the target user from a content access history that does not include the user's registration information is used as potential user data. Using the latent data extractor to be output and the relevant item specified as a condition for associating with the record of the potential user data, the data set storing the access history different from the access history related to the record of the potential user data and a potential user data appending unit for appending records to records of potential user data. [Selection diagram] Fig. 1

Description

The present invention relates to a technique for extracting latent user data from an access history of content that does not include user registration information and adding relevant data of another dataset group to the latent user data.

In the field of marketing, there is known a technique of acquiring information of a target user from a plurality of databases by using information such as attributes and behavior history of users who use the content. As a technique for acquiring user information, for example, Japanese Patent Application Laid-Open No. 1 and Patent Document 2 are known.

Japanese Patent Application Laid-Open No. 1 describes a customer data collection means for collecting customer data including customer information of each company in a marketing data providing system for providing marketing data to a plurality of companies, and marketing for identifying the provided customer data. The data provision system is disclosed.

Patent Document 2 is characterized in that it includes a supplementary unit that estimates the characteristics of the user using the attributes of the user and supplements the characteristic information of the target user based on the characteristic information of a similar user similar to the target user. The information processing device is disclosed.

Japanese Patent Application Laid-Open No. 2003-288464 Japanese Unexamined Patent Publication No. 2020-035219

A contributor who provides content information to a website may intend to create a new business depending on the browsing user who accesses the provided content. Some of these browsing users have similar tastes and similarities to, for example, other than the targetable users who register on the website to download materials and provide personal attribute information. However, there are potential users who only access the website and do not provide personal attribute information. The potential user refers to a user whose content access history is similar to that of the target user, and the target user refers to a user who wants to use his / her own product service among the targetable users. If information that can be used for marketing is obtained for this potential user in addition to the information of the target user that was previously available, it will be possible to create a new business. Further, the information that can be used for marketing includes not only information that can directly identify an individual user and sell a product or service, but also other information such as a search keyword and purchase information, such as a user's preference tendency. Therefore, even a potential user who does not include the registration information of the user who uses the content of the website can be information that can be utilized for marketing by adding data related to this potential user from another data set group. ..

In Japanese Patent Laid-Open No. 1 of the above-mentioned conventional example, data is added by name identification to data that can be identified by a target individual, but it is considered to add data when an individual cannot be identified. There was a problem that it wasn't there.

In Patent Document 2 of the above-mentioned conventional example, in order to be used for optimizing one's own service, information on a user who can be identified by a specific target individual is used as an analysis result of another service to obtain a target user or a similar user. It was supplemented using information, not for similar users who could not be identified by an individual.

Therefore, the present invention has been made in view of the above problems, and is a data addition system capable of adding data that can be used for marketing to a potential user who cannot identify an individual similar to a target user, and data addition. The purpose is to provide a method.

The data addition system according to the present invention is data including a record of a potential user who is similar to the target user including a specified condition by a certain amount or more and a record of the target user from the access history of the content which does not include the user registration information. The access history related to the record of the latent user data by using the latent data extraction unit that outputs the latent user data and the related item specified as the condition for associating the record of the latent user data with the data. It is configured as a data addition system characterized by having a latent user data addition unit that adds a record of a data set storing different access histories to the record of the latent user data.

According to the present invention, it is possible to add data that can be used for marketing to a potential user who cannot identify an individual similar to the target user.

It is a block diagram which shows an example of the structure of the latent user data addition system in this Example. It is a figure which shows an example of the flowchart which shows the processing procedure of the processing performed by the latent user data extraction unit shown in FIG. It is a figure which shows an example of the flowchart which shows the processing procedure of the processing performed by the related data analogy part shown in FIG. It is a figure which shows an example of the flowchart which shows the processing procedure of the processing performed by the data addition part shown in FIG. 1. It is a figure which shows an example of the data of the access history shown in FIG. It is a figure which shows an example of the data set of the data set group shown in FIG. It is a figure which shows an example of the input data of the input part shown in FIG. It is a figure which shows an example of the processing result of the latent user data extraction part shown in FIG. It is a figure which shows an example of the relationship between the access history and a target user in this Example. It is a figure which shows an example of the relationship between a target user and a potential user in this Example. It is a figure which shows an example of the relationship between the access history and the data set 1 in this Example. It is a figure which shows an example of the relationship between the access history and the data set 2 in this Example. It is a figure which shows an example of the relationship between the access history and the data set 3 in this Example. It is a figure which shows an example of the output result in this Example. It is a computer schematic diagram of a system.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The following description and drawings are examples for explaining the present invention, and are appropriately omitted and simplified for the sake of clarification of the description. The present invention can also be implemented in various other forms. Unless otherwise specified, each component may be singular or plural.

The positions, sizes, shapes, ranges, etc. of each component shown in the drawings may not represent the actual positions, sizes, shapes, ranges, etc., in order to facilitate understanding of the invention. Therefore, the present invention is not necessarily limited to the position, size, shape, range and the like disclosed in the drawings.

In the following description, various information may be described by expressions such as "table" and "list", but various information may be expressed by a data structure other than these. The "XX table", "XX list", etc. may be referred to as "XX information" to show that they do not depend on the data structure. When expressions such as "identification information", "identifier", "name", "ID", and "number" are used in explaining the identification information, these can be replaced with each other.

When there are a plurality of components having the same or similar functions, they may be described by adding different subscripts to the same reference numerals. However, if it is not necessary to distinguish between these plurality of components, the subscripts may be omitted for explanation.

Further, in the following description, a process performed by executing a program may be described, but the program is determined by being executed by a processor (for example, a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit)). In order to perform the processed processing while appropriately using storage resources (for example, memory) and / or interface devices (for example, communication port), the main body of the processing may be a processor. Similarly, the subject of the process of executing the program may be a controller, an apparatus, a system, a computer, or a node having a processor. The main body of the processing performed by executing the program may be any arithmetic unit, and may include a dedicated circuit (for example, FPGA (Field-Programmable Gate Array) or ASIC (Application Specific Integrated Circuit)) for performing specific processing. ..

The program may be installed from a program source into a device such as a calculator. The program source may be, for example, a program distribution server or a computer-readable storage medium. When the program source is a program distribution server, the program distribution server includes a processor and a storage resource for storing the program to be distributed, and the processor of the program distribution server may distribute the program to be distributed to other computers. Further, in the following description, two or more programs may be realized as one program, or one program may be realized as two or more programs.

FIG. 1 is a block diagram showing an example of the configuration of the latent user data addition system 1000 in this embodiment. The latent user data addition system 1000 includes an input unit 001, a content access history 002, a latent user data extraction unit 003, a data set group 004 having a plurality of data sets, a related data estimation unit 0051, and a data acquisition unit. It has a latent user data addition unit 005 including 0052. These specific functions will be described later using a flowchart.

The latent user data addition system 1000 includes, for example, a CPU 1501, a memory 1502, an external storage device 1503 such as an HDD (Hard Disk Drive), a CD (Compact Disk), or a USB, as shown in FIG. 15 (computer schematic diagram). A reading device 1507 that reads and writes information to and from a portable storage medium 1508 such as a memory, an input device 1506 such as a keyboard and a mouse, an output device 1505 such as a display, and a NIC (Network) for connecting to a communication network. It can be realized by a general computer 1500 including a communication device 1504 such as an Interface Card) and an internal communication line (referred to as a system bus) 1509 such as a system bus connecting them.

Various data stored in each system or device or used for processing (for example, the production-related data 2 described above) can be realized by the CPU 1501 reading from the memory 1502 or the external storage device 1503 and using the data. Further, each functional unit (described later) possessed by each system or device can be realized by the CPU 1501 loading a predetermined program stored in the external storage device 1503 into the memory 1502 and executing the program.

The predetermined program described above is stored (downloaded) in the external storage device 1503 from the storage medium 1508 via the reading device 1507 or from the network via the communication device 1504, and then loaded onto the memory 1502. It may be executed by the CPU 1501. It may also be loaded directly onto the memory 1502 from the storage medium 1508 via the reading device 1507 or from the network via the communication device 1504 and executed by the CPU 1501.

In the following, the case where the latent user data addition system 1000 is configured by one computer is illustrated, but these functional units may be configured by a plurality of computers. Further, all or a part of these functional units may be distributed in one or a plurality of computers such as a cloud, and the same functions may be realized by communicating with each other via a network. ..

FIG. 2 is a diagram showing an example of a flowchart showing a processing procedure of processing performed by the latent user data extraction unit shown in FIG. 1. The latent user data extraction unit 003 calculates the similarity between the user of each record in the access history and the target user (step 102), and the similarity between the user of each record in the access history and the target user data is equal to or higher than a threshold value α. Data is extracted (step 103). The target user is a user who wants to use his / her own product / service among the targetable users. For example, as described later with reference to FIG. 7, a condition to be a target user is specified from the input unit 001. To. As a method of calculating the similarity, for example, there is a method in which the latent user data extraction unit 003 clusters the access history in advance and calculates the distance to the target user as the similarity. However, the similarity calculation method is not limited to this. The processing result of the latent user data extraction unit 003 will be described later with reference to FIG.

FIG. 3 is a diagram showing an example of a flowchart showing a processing procedure of processing performed by the related data analogy unit shown in FIG. 1. The related data analogy unit 0051 selects a data set from the data set group (step 202), refers to the related item specified in the input unit 001 and the processing method, and determines whether there is an item for which name identification processing is specified in the processing method. Determination (step 203). When the related data analogy section 0051 determines that there is an item for which the name identification process is specified in the processing method (step 203; YES), the name identification process is performed (step 204). Further, the related data analogy unit 0051 determines whether or not there is an item for which the relevance calculation is specified in the processing method (step 205). When the related data analogy section 0051 determines that the processing method has an item for which the relevance calculation is specified (step 205; YES), the relevance calculation unit 0051 calculates the relevance degree (step 206). When the related data analogy unit 0051 determines whether or not all the data sets have been processed in the data set group (step 207) and determines that all the data sets have been processed (step 207; YES). The process shown in FIG. 3 is terminated. On the other hand, when the related data analogy unit 0051 determines that all the data sets have not been processed (step 207; NO), the process returns to the selection of the data set in step 202. The degree of relevance in step 206 is calculated by using a method for measuring the degree of similarity between words by natural language processing. However, the relevance calculation method is not limited to this.

FIG. 4 is a diagram showing an example of a flowchart showing a processing procedure of processing performed by the data addition unit shown in FIG. 1. The data addition unit 0052 adds the result of the record subjected to the name identification processing by the related data analogy unit 0051 to the latent user data (step 302), and also adds the relational result of the item whose relevance degree has been calculated and the relevance degree β (step). 303).

FIG. 5 is a diagram showing an example of the access history data shown in FIG. The access history 002 shown in FIG. 1 is information indicating a history of access to a website (for example, a blog) that can be accessed by an unspecified user without requiring registration. The content access history A401, which is an example of the access history 002, is collected by, for example, a web server at a predetermined cycle or the like.

As shown in FIG. 5A, the access history A401 is a table including the item group A-1 {IP address, access time, number of visits, visited page, browsing tag, search keyword} in one record. "No" is a serial number for identifying the record, and in (a) of FIG. 5, the record from No. 1 to _N0 is recorded as the access history A401. The access history A401 is stored for each content. Among the access users who have accessed this content, the user who is interested in the content has the item group A- as shown in FIG. 5 (b) in order to download the material in order to obtain more detailed information. 2 It is assumed that the material is downloaded from the above website by sending {name, affiliation, purpose of investigation, download material link} to the web server. The web server that has collected the access history A401 uses the data having this item group A-2 as the additional acquisition data, and as a result of the addition, the access history + the additional acquisition data (additional acquisition data access history) A402 is configured. In FIG. 5B, the records of Nos. 1, 3, 4, and _N0 are the access histories of the user who accessed the contents of the website and downloaded the materials, and the records of Nos. 2 and 4 are the contents of the website. It shows that it is the access history of the user who accessed the site but did not download the material. Similar to the access history A401, the additional acquisition data access history A402 is also collected by, for example, a web server at a predetermined cycle or the like.

FIG. 6 is a diagram showing an example of a data set of the data set group shown in FIG. The data set group 004 shown in FIG. 1 shows a history of accessing a website different from the website stored as the access history 002 (for example, an operation site that provides a blog environment or another operation site). Information. In the following, three types of data sets are assumed as the data sets included in the data set group.

The first is the data set B501 (data set 1) in which the user group assumed for the contents of the website operated by the user is the same as the user group assumed for the accessed contents stored in the access history. be. The data set B501 includes the item group B-1 {IP, access time, number of visits, visited page, search keyword}. “No” is a serial number for identifying the records, and in FIG. 6A, the records from No. ₁ to N1 are recorded as the data set B501. The data set B501 may be stored for each content.

The second is that the user group assumed for the content of the website operated by oneself is the same as the user group expected for the accessed content stored in the access history, but the content of the website operated by oneself Data set B502 (data set 2) that requires registration. Data set B502 includes item group B-2 {name, affiliation, IP, charge, purchase information}. “No” is a serial number for identifying records, and in FIG. 6B, records 1 to _N2 are recorded as data set B502. The data set B502 may be stored for each content.

The third is the data set B503 (data set 3), which is operated by another person and it is unknown whether or not the user layer similar to the user layer assumed in the content of the website operated by the other person is included. It is assumed that the data set B503 includes the item group B-3 {name, affiliation, posted article content}. The data set B503 may be stored for each content.

FIG. 7 is a diagram showing an example of input data of the input unit shown in FIG. 1. The input unit 001 accepts the input of the data set of the additional acquisition data access history A402 having the additional acquisition data as the data of the access history 002. Further, as the information about the target user, an example of an access user whose {investigation purpose} of the item group A-2 of the access history 002 corresponds to "AI application" is input. Further, for each data set, the related items for the access history 002 and the data in which the processing method is input are input. Information about the target user, related items for the access history 002, and the processing method shall be determined in advance by the user of this system. The designation of related items includes not only one-to-one designation but also one-to-many, many-to-one, and many-to-many designations.

In FIG. 7, for example, for the data set 1, the input unit 001 accepts the designation of {IP address}, {visit page}, and {search keyword} as related items for the access history 002, and for items other than these, the input unit 001 accepts the designation. Indicates that it was not specified as a related item. Further, the input unit 001 indicates that the designation of {name identification}, {relevance degree calculation}, and {relevance degree calculation} has been accepted as the processing method of these items designated as the related items, respectively. It can be seen that these items are specified in the same manner for the data set 2 and the data set 3.

FIG. 8 is a diagram showing an example of the processing result of the latent user data extraction unit shown in FIG. The latent user data extraction unit 003 calculates the similarity between the user stored in the access history and the target user in the similarity calculation in step 102 shown in FIG. 2, and in step 103, the user and the target user in the access history. Extract data with a threshold α or higher that has a degree of similarity with.

In FIG. 8, the similarity between the user stored in the access history and the target user is calculated by using the {visit count, visited page, browsing tag} of the item group A-1 as a key, and the access user having a high degree of similarity is calculated. It is extracted as latent user data 701.

More specifically, in FIG. 8, the latent user data extraction unit 003 is previously included in the additional acquisition data access history A402 among the records included in the additional acquisition data access history A402 shown in FIG. 5 (b). The {visit count, visited page, browsing tag} of the item group A-1 specified as a similar condition for extracting records that are similar to a certain level or more are the same (or include at least a part, and the content of the item is semantically constant. Extract the data (similar to the above). Further, the latent user data extraction unit 003 is located in the {visit count, visit page, browsing tag} of the item group A-1 included in the record for which the condition is specified as the target user in the extracted data. Records containing {visit count, visit page, browsing tag} of the above item group A-1 having a threshold value of α or more are extracted. Thereby, in the additional acquisition data access history A402, the data including the record of the target user and the record of the potential user having a certain degree or more similar to the target user can be extracted as the potential user data 701. As this similarity calculation method, for example, there is a method in which access history 002 is clustered in advance and the closeness of the distance to the cluster of the target user is calculated as the similarity. However, the similarity calculation method is not limited to this. In FIG. 8, among the records of the latent user data 701, the user stored as the records of Nos. 1 and 3 uses the {investigation purpose} of the item group A-2 of the access history 002 as the "AI application" of the website. It indicates that the user has accessed and input the content and downloaded the material, that is, the target user who meets the conditions input by the input unit 001 in (2) of FIG. Further, the users (Nos. 5, 7, and 11) other than the target users stored as the records of Nos. 1 and 3 indicate that they are potential users who are similar to the target users by a certain amount or more.

FIG. 9 is a diagram showing an example of the relationship between the access history 002 and the target user in this embodiment. In FIG. 9, among the access users, the access user U1 having only the item group A-1 and the additional acquisition user who has downloaded the material and acquired the additional acquisition data having the item group A-2. U2 (for example, the names "Tanaka" and "Ito" in (b) of FIG. 5) and the target user U3 (for example, item group A-2 {for example, which specifies the target range to be acquired among the additional acquisition users. The purpose of the survey} indicates that there are names "Sato" and "Suzuki") whose "AI application" is used. Here, the access user U1 includes the additional acquisition user U2, and the additional acquisition user U2 includes the target user U3. If the target user can be specified using the item group A-1, the additional acquisition user data is not always required.

FIG. 10 is a diagram showing an example of the relationship between the target user and the potential user in this embodiment. FIG. 10 shows that among the access users, there is a latent user U4 having a similarity α or higher with the target user U3 in a certain classification X. This classification X refers to {visit count, visit page, browsing tag} in the item group A-1 in FIG. 8, and the item is not limited to one, and is dimensionally compressed by principal component analysis or the like. Is also a target. The target user is included in the potential user.

FIG. 11 is a diagram showing an example of the relationship between the access history 002 and the data set 1 in this embodiment. The processing result when the data set 1 is selected in the processing (step 202) for selecting the data set in the related data analogy section 0051 is shown. From FIG. 7, since the related item specified in the name identification process (step 204) includes the item of the data set 1 = "IP address", the related data analogy section 0051 is based on the item of the access history = "IP address". , Perform name identification processing (step 204) for the data set 1. Here, even if the IP address is the same, the user may be different. However, in the data set 1, in the system that manages the contents of the website operated by itself, for example, due to the structure of the system, the IP addresses of different users (for example, users having a plurality of handle names) are paired. If there is a premise associated with 1, it is treated as an item of name identification processing (step 204). Further, for example, the item of access history = "access time" and the item of data set 1 = "access time" must be a data set related to a system in which data set 1 accesses at the same time even if the names are the same. For example, it is not specified as a related item. However, even if the access history and the access time for the data set 1 are not necessarily the same time, for example, a certain time width is provided in consideration of the delay due to the communication environment, and the time width includes the above-mentioned access time. May be treated as an item of the above-mentioned name identification processing.

In any case, since it is the user of this system who can determine which items are to be linked in this way, it is necessary to specify the related items and the processing method in the input unit 001. In (1) name identification processing of FIG. 11, the user data U41 of one potential user U4 included in the access history 002 is the same as the user data U11 after name identification among the access users U1 included in the data set 1. Indicates that the user has been determined. Further, from FIG. 7, the designated items include the item of the data set 1 = "visit page" and "search keyword". Therefore, in the relevance calculation (step 206), the related data analogy unit 0051 adds the access history item = "visit page" and the "search keyword" designated for each to the access history 002. The relevance calculation (step 206) between the included potential user or target user and the access user included in the data set 1 is executed.

In (2) relevance calculation of FIG. 11, one user data U31 of the target user U3 included in the access history 002 is combined with the user data U12 which is one of the user data of the access user U1 included in the data set 1. It shows that they were associated with the degree of relevance β ₁₁ . Further, the user data U42 of one potential user U4 included in the access history 002 is the user data U13 and U14 which are two of the user data of the access users U1 included in the data set 1, and the degree of association β ₁₂ respectively. It is shown that it was linked with β ₁₃ . In (2) of FIG. 11, the case of the “visit page” is illustrated as the item of the data set 1, but the “search keyword” can be considered in the same manner.

FIG. 12 is a diagram showing an example of the relationship between the access history 002 and the data set 2 in this embodiment. The processing result when the data set 2 is selected in the processing (step 202) for selecting the data set in the related data analogy section 0051 is shown. From FIG. 7, since the related item specified in the name identification process (step 204) includes the item of the data set 2 = "name" AND "affiliation", the related data analogy section 0051 has the item of the access history = "name". By "AND" belonging ", the name identification process (step 204) is performed on the data set 2. In (1) name identification processing of FIG. 12, one user data U32 of the target user U3 among the potential users U4 included in the access history 002 is the user data after name identification among the registered users U5 included in the data set 2. It is shown that the user is determined to be the same user in association with the user data U51, which is one of the above.

Further, from FIG. 7, the designated item includes the item of the data set 2 = "in charge". This item has a different name from the access history item = "investigation purpose", but if the structure of the system allows the use of different data sets, similar content may be used even if the item name is different. You may have entered it. Therefore, as shown in FIG. 7, it is possible for the user to arbitrarily specify an item including similar contents in the input unit 001. In addition to manually inputting this item specification, for example, the contents of these two items are compared with each other, and one of the items contains similar contents with the contents of another item and a certain threshold value or more. May be determined and automatically specified as a related item as the default value.

In (2) relevance calculation of FIG. 12, user data U52 and U53 in which one user data U33 of the target user U3 included in the access history 002 is two user data of the registered users U5 included in the data set 2 are used. And, respectively, it is shown that they are linked by the degree of association β ₂₁ and β ₂₂ . In the relevance calculation (step 206), since the access history including the item group A-2 is targeted, only the target user having the item group A-2 is processed.

Further, although not shown in the figure, the data set 2 is the data registered in the system of the website operated by itself, and the data of the system used by the same user layer as the user layer of the access history 002. Consider the case where. In this case, if there is a related item specified in the item group A-1 and there is a user whose relationship between the data set 2 and the access history 002 is higher than a certain threshold value for the related item, the individual cannot be identified. An individual may be identified by adding the registered user data of a user whose relevance is higher than a certain threshold value among the registered user data U5 of the data set 2 to the user data of the latent user.

FIG. 13 is a diagram showing an example of the relationship between the access history 002 and the data set 3 in this embodiment. The processing result when the data set 3 is selected in the processing (step 202) for selecting the data set in the related data analogy section 0051 is shown. From FIG. 7, since the related item specified in the name identification process (step 204) includes the item of the data set 1 = "name" AND "affiliation", the related data analogy section 0051 has the item of the access history = "name". By "AND" belonging ", the name identification process (step 204) is performed on the data set 3.

In (1) name identification processing of FIG. 13, one user data U34 of the target user U3 in the potential user U4 included in the access history 002 is one of the registered users U5 included in the data set 3 after the name identification. It shows that it is determined as the same user in association with the user data U61 which is the user data. Further, from FIG. 7, the designated item includes the item of the data set 3 = "posted article content". However, the item may be included in the registration data registered by a large number of users such as SNS (Social Networking Service) in a system that manages the contents of a website operated by another person. In such a case, if a large number of similar users are linked, the degree of relevance is calculated, and data is added, for example, by linking the item so that the item "posted article content" can be accessed via the item. It can be complicated. Therefore, in (2) of FIG. 13, it is assumed that the groups are linked to each other as surrounded by the alternate long and short dash line of the data set 3 to output the characteristics of the highly related groups. As for the method of grouping, the related data analogy unit 0051 can cluster the data of the data set 3 in advance to classify the data, but the method is not limited to this. In (2) relevance calculation of FIG. 13, the two user data U43 and U44 of the potential user U4 included in the access history 002 are each of the registered users included in the data set 3 and have the above-mentioned classified relevance. It is shown that the user data of the high groups G61 and G62 are associated with the relevance degree β ₃₁ and β ₃₂ , respectively.

FIG. 14 is a diagram showing an example of an output result in this embodiment. FIG. 14 illustrates the additional data added to the item group A-1 and the item group A-2 shown in FIG. 5, but in reality, each record of the additional data includes the item group A-. 1 and item group A-2 are associated with each other. Further, FIG. 14 shows an output example 1301 when the data addition unit 0052 shown in FIG. 4 executes processing on the latent user data as the additional data.

The related data analogy section 0051 adds the record of the name identification process (step 204) for each data set to the item group A-1 and the item group A-2 shown in FIG. 5 as additional data. Further, the related data analogy unit 0051 outputs the related result and the related degree β for the related item used for the related degree calculation (step 206) for each data set. Although the related result is an example of outputting the contents of the related item here, the record result may be output in the same manner as the name identification process.

As a result, as shown in FIG. 14, in addition to the record of the target user initially designated from the input unit 001 (for example, the record of No. 1 or No. 3), the potential user other than the target user included in the potential user. It is possible to add related data to a record (for example, a record of No. 7 or No. 11). By adding related data to the record of the potential user, for example, No. The potential user of 7 is interested in material A and material C from FIG. 8, and considers AI application from FIG. 14 with a degree of relevance β = 0.8 (that is, a degree of relevance of 80%) from the content of the posted article. It turns out that there is a possibility of being a graduate student, and it is possible to envision a field of interest to be marketed. In addition, No. The 11 potential users are interested in the material A from FIG. 8, and from the content of the posted article, the relevance β = 0.8 (that is, the relevance of 80%), and from FIG. It may be possible to find out that there is a possibility of working workers and to anticipate the needs of the site. Therefore, it can be said that this system can acquire information that can be used for marketing by adding data related to potential user data.

More specifically, in the latent user data addition system 1000 in this embodiment, the specified conditions (for example, the item group A- of the access history 002 in FIG. 7) are specified from the access history 002 of the content that does not include the user registration information. A record of a potential user (for example, a record of records Nos. 5, 7 and 11 in FIG. 8) and a record of the target user that are similar to the target user by a certain amount or more including the condition that the {investigation purpose} of 2 is "AI application"). Related items specified as conditions for associating the data including (for example, the records of records No. 1 and 3 in FIG. 8) with the latent data extraction unit 003 that outputs the data as latent user data and the record of the latent user data. For example, using {IP address}, {visit page}, {search keyword} in FIG. 7, a data set (for example,) that stores an access history different from the access history related to the record of the latent user data (for example). It has a latent user data addition unit 005 that adds the records of the data sets 1 to 3) of FIG. 6 to the record of the latent user data. Therefore, it is possible to add data that can be used for marketing from various data other than the access history for potential users whose individuals similar to the target user included in the access history cannot be identified.

Further, the latent user data addition unit 005 may use the records of the data set and the latent user based on the contents of the related items (for example, the contents of the {visit page}, the character string or the value of the {search keyword}). It has a related data estimation unit 0051 that calculates the degree of relevance of the data to the record. Therefore, it is possible to grasp the degree of relevance between various records of the data set and various records of the latent user data.

Further, the latent user data addition unit 005 has a data addition unit 0052 that outputs output result data in which the record of the latent user data for which the degree of relevance is calculated and the record of the data set are associated with each other. Therefore, the record of the latent user data for which the degree of relevance has been calculated can be associated with the record of the data set and presented to the user.

Further, the latent data extraction unit 003 {visits number of times of the item group A-1 designated as a similar condition for extracting records having a certain degree or more similar in a predetermined item (for example, in FIG. 8 in FIG. 8) from the access history. , Visit page, browsing tag}) is extracted from the records of users whose contents are similar to a certain level or more, and among the extracted records, the records of potential users and the records of the target user that are similar to the target user by a certain amount or more are extracted. The included data is output, for example, as latent user data shown in FIG. Therefore, for example, after the user of this system specifies a desired item group, data including a record of a potential user and a record of the target user that are similar to the target user by a certain amount or more is extracted and output as potential user data. Can be done.

Further, when the name identification process is specified for the related item for each data set of the data set group, the related data guessing unit 0051 performs the name identification process for the record of the latent user data, and the related item. When it is specified to calculate the degree of relevance, the degree of relevance can be calculated for the records of the above latent user data set.

Further, the data addition unit 0052 is used to calculate the relevance degree (for example, β = 0.8 of record No. 7 in FIG. 14) calculated by the related data analogy unit 0051 and the relevance degree. The output result data including the contents of related items (for example, “clustering” of record No. 7 in FIG. 14) is output to a display unit (for example, an output device 1505 such as a display). Therefore, the user of this system can grasp the details of the degree of relevance, such as what kind of content is based on and how much relevance is, at a glance.

As described above, by using the access history of the content that does not include the user registration information, the information of the target user that is an example of the acquisition target, the related items with each data set, and the processing method are input, so that the potential user can be used. It is possible to extract data and add data related to latent user data from another dataset group. As a result, in addition to the information that can directly identify the individual user and sell the product / service, the information that can be used for marting such as the preference tendency can be acquired, and it is possible to create a new business.

The present invention is not limited to the above-described embodiment, and includes various modifications. For example, the above-described embodiment is described in detail in order to explain the present invention in an easy-to-understand manner, and is not necessarily limited to the one including all the configurations described. Further, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Further, for a part of the configuration of each embodiment, any of addition, deletion, or replacement of other configurations can be applied alone or in combination.

1000: Latent user data addition system 001: Input unit 002: Access history 003: Latent user data extraction unit 004: Data set group 005: Latent user data addition unit 0051: Related data analogy unit 0052: Data addition unit

Claims (12)

From the access history of the content that does not include the user registration information, the data including the record of the latent user that is similar to the target user including the specified condition by a certain amount or more and the record of the target user is output as the latent user data. Data extraction unit and
Using the related item specified as a condition for associating with the record of the latent user data, the record of the data set related to the record of the latent user data and storing the access history different from the access history is described as the latent. The latent user data addition part to be added to the user data record,
A data addition system characterized by having. The data addition system according to claim 1.
The latent user data addition unit is a related data analogy unit that calculates the degree of relevance between the record of the data set and the record of the latent user data based on the contents of the related item.
A data addition system characterized by having. The data addition system according to claim 2.
The latent user data addition unit is a data addition unit that outputs output result data in which a record of the latent user data for which the degree of relevance is calculated and a record of the data set are associated with each other.
A data addition system characterized by having. The data addition system according to claim 1.
The latent data extraction unit extracts a record of a user having similar contents of a predetermined item from the access history, and among the extracted records, a record of a latent user having a certain degree or more similar to the target user and the target. Output data including user records as latent user data,
A data addition system characterized by this. The data addition system according to claim 2.
When the name identification process is specified for the related item for each data set of the data set group, the related data guessing unit performs the name identification process for the record of the latent user data and calculates the degree of relevance for the related item. If it is specified, the degree of relevance is calculated for the records in the latent user data set.
A data addition system characterized by this. The data addition system according to claim 3.
The data addition unit outputs the output result data including the relevance degree calculated by the related data analogy unit and the contents of the related item used for calculating the relevance degree to the display unit.
A data addition system characterized by this. It is a data addition method performed by a computer.
From the access history of the content that does not include the user's registration information, the latent data extraction unit latently collects data including the record of the latent user that is similar to the target user including the specified condition by a certain amount or more and the record of the target user. Output as user data
Data in which the latent user data addition unit stores an access history different from the access history related to the record of the latent user data by using the related item specified as a condition for associating with the record of the latent user data. Add the record of the set to the record of the latent user data.
A data addition method characterized by this. The data addition method according to claim 7.
The related data analogy unit of the latent user data addition unit calculates the degree of association between the record of the data set and the record of the latent user data based on the contents of the related item.
A data addition method characterized by this. The data addition method according to claim 8.
The data addition unit of the latent user data addition unit outputs output result data in which the record of the latent user data for which the degree of relevance is calculated and the record of the data set are associated with each other.
A data addition method characterized by this. The data addition method according to claim 7.
The latent data extraction unit extracts a record of a user having similar contents of a predetermined item from the access history, and among the extracted records, a record of a latent user having a certain degree or more similar to the target user and the target. Output data including user records as latent user data,
A data addition method characterized by this. The data addition method according to claim 8.
When the related data guessing unit specifies the name identification process for the related item for each data set of the data set group, the name identification process is performed on the record of the latent user data, and the degree of relevance is calculated for the related item. If it is specified, the degree of relevance is calculated for the records in the latent user data set.
A data addition method characterized by this. The data addition method according to claim 9.
The data addition unit outputs the output result data including the relevance degree calculated by the related data analogy unit and the contents of the related item used for calculating the relevance degree to the display unit.
A data addition method characterized by this.

Images