JP5034279B2 - Attribute estimation program and attribute information providing system - Google Patents

Description

  The present invention relates to an attribute estimation program and an attribute information providing system for estimating an attribute of a user of a terminal device.

  Traditionally, for companies that provide products and services, collect and use personal information (attribute information) about who the consumer is in order to change the method of providing the product or service according to the consumer. Things have been done. Such collection and use of personal information is not limited to face-to-face sales and advertisements, but is also performed through online sales using the Internet or homepage searches. For example, URL information of websites browsed by users can be obtained. A technique for analyzing and estimating a user attribute using the technique is disclosed (for example, Patent Document 1, Patent Document 2, and the like).

JP 2003-16345 A Special table 2003-529127 gazette

  By the way, in the above-described conventional techniques (Patent Document 1 and Patent Document 2), even when the user browses a website different from his / her purpose, the user's attribute is analyzed, and thus the estimated noise is generated. In many cases, there is a problem that the accuracy of the estimation result obtained by estimating the attribute of the user of the terminal device is not high.

  Accordingly, the present invention has been made to solve the above-described problems of the prior art, and provides an attribute estimation program and an attribute information providing system capable of systematically estimating the attributes of a user of a terminal device The purpose is to do.

In order to solve the above-described problems and achieve the object, the present invention provides an attribute estimation program for causing a computer to execute a method for estimating the attribute of a user of a terminal device, which is input from the terminal device by the user or A search word storing procedure in which a search word to be selected and a user ID for uniquely identifying the user are associated with each other and stored in a search word storage unit, and a user having a predetermined attribute and the attribute is connected to the terminal Attribute storage procedure for associating search words that are highly likely to be input or selected from the apparatus and storing them in the attribute storage means, search words stored in the search word storage means, and search words stored in the attribute storage means Based on the match, the attribute estimation procedure for estimating the user attribute, the user attribute estimated by the attribute estimation procedure, and the user ID are associated with each other. Characterized in that to execute the user attribute storage procedure of storing the unit and the computer.

Further, the present invention is the above invention, wherein the attribute storing procedure newly stores a search word and an attribute used for estimation by the attribute estimation procedure in association with each other in the attribute storage unit, and the attribute estimation procedure includes: The computer is caused to estimate the attribute of the user by using a search word and an attribute newly stored in the search word storage means.

In the above invention, the present invention further includes estimated accuracy calculation means for calculating an estimation accuracy indicating the accuracy of the attribute estimated by the attribute estimation procedure, and the user attribute storage procedure includes the estimation accuracy calculation The computer is caused to further associate the estimated accuracy calculated by the means and store it in the user attribute storage means.

According to another aspect of the present invention, a terminal device used by a user, a web server device that provides information to the terminal device of the user, and an information providing server device that estimates the attribute of the user are connected to a network. In the information providing system configured, the information providing server device associates a search term input or selected from the terminal device by the user with a user ID for uniquely identifying the user. Search word storage means for storing the search attribute, attribute storage means for storing a predetermined attribute and a search word that a user having the attribute is likely to input or select from the terminal device, and the search word storage Attribute estimation means for estimating the attribute of the user based on a match between the search word stored in the means and the search word stored in the attribute storage means, and estimated by the attribute estimation means User attribute storage means for storing the user attribute and the user ID in association with each other, and when the request related to the user attribute is received from the web server device, the attribute estimation means estimates the user attribute. Attribute transmitting means for transmitting the attribute of the user to the web server device, and the web server device transmits a request relating to the attribute of the user of the terminal device to the information providing server device. Attribute receiving means for receiving an attribute from the information providing server apparatus, and information transmitting means for transmitting predetermined information to the terminal apparatus based on the attribute of the user received by the attribute receiving means. It is characterized by.

According to another aspect of the present invention, a terminal device used by a user, a web server device that provides information to the terminal device of the user, and an information providing server device that estimates the attribute of the user are connected to a network. In the information providing system configured, the information providing server device includes a user ID for uniquely identifying a search term input or selected from the terminal device by the user from the terminal device and the user. Search word storage means for storing the above in association with each other, attribute storage means for storing a predetermined attribute and a search word that is likely to be input or selected by a user having the attribute, and the search word storage Attribute estimation means for estimating the attribute of the user based on a match between the search word stored in the means and the search word stored in the attribute storage means, and estimated by the attribute estimation means When a user attribute storage unit that stores the user attribute and the user ID in association with each other and a request related to the user attribute from the terminal device that is accessing the web server device are received. An information transmitting unit that transmits predetermined information corresponding to the attribute of the user estimated by the attribute estimating unit to the terminal device, and the web server device receives the access from the terminal device and receives the access response And an access response means for performing.

According to the present invention, the search term input or selected by the user from the terminal device and the user ID for uniquely identifying the user are stored in the search term storage unit in association with each other, A search term that is highly likely to be input or selected from a terminal device by a user having an attribute is stored in the attribute storage means, and a search word stored in the search word storage means and a search stored in the attribute storage means Since the user attribute is estimated based on the match with the word and the estimated user attribute and the user ID are stored in association with each other, noise is hardly included as a result of reflecting the user's intention. Attributes (age group, gender, etc.) from search terms, such as words that represent the same thing but are different in gender (for example, nurses and nurses) and words used by specific age groups (for example, hiring new graduates) , Residential area, etc.) It may result, it is possible to accurately estimate the attribute of the user of the terminal device.

According to the present invention, the search terms and attributes used for estimation are newly stored in association with each other, and the user attributes are estimated using the search terms and attributes newly stored in the search term storage means. Therefore, by adding search terms and attributes used for user attribute estimation, it is possible to improve the accuracy of estimation results as a result of improving the coverage of predetermined attributes.

Also, according to the present invention, the estimation accuracy indicating the accuracy of the estimated attribute is calculated, and the calculated estimation accuracy is further stored in association with each other. Therefore, the estimation accuracy is calculated in association with the attribute and the user ID. As a result, the accuracy of the estimation result of each user ID or attribute can be known, and it can be used as a guideline for further improving the accuracy of the estimation result.

Further, according to the present invention, the information providing server device stores a search term input or selected by the user from the terminal device and a user ID for uniquely identifying the user in association with each other. A search word stored in the search word storage unit and a search word stored in the attribute storage unit are stored in association with the search term that is highly likely to be input or selected from the terminal device by the user having the attribute. When the user attribute is estimated, the estimated user attribute and the user ID are stored in association with each other, and a request related to the user attribute is received from the web server device. The estimated user attribute is transmitted to the web server apparatus, and the web server apparatus transmits a request related to the user attribute of the terminal apparatus to the information providing server apparatus, and the user attribute is transmitted from the information providing server apparatus. Receive, receive Since the predetermined information is transmitted to the terminal device based on the attribute of the user, even when the web server device and the information providing server device belong to different domains (for example, the search service and the attribute estimation service are performed by different companies) Case), the user ID is not known to the web server device, and it is possible to perform a service using the attribute information estimated by the information providing server device.

Further, according to the present invention, the information providing server device stores a search term input or selected from the terminal device by the user from the terminal device and a user ID for uniquely identifying the user. , A predetermined attribute and a search word that is likely to be input or selected by a user having the attribute are stored in association with each other, and based on a match between the stored search word and the search word stored in the attribute storage unit The user attribute is estimated, the estimated user attribute and the user ID are stored in association with each other, and a request relating to the user attribute is received from the terminal device accessing the web server device In this case, predetermined information corresponding to the estimated user attribute is transmitted to the terminal device, and the web server device accepts access from the terminal device and makes an access response. When the server device belongs to the same domain (for example, when the search service and the advertisement service are performed by the same company), it is possible to reduce the cost when constructing the system without providing a complicated system configuration and to provide information The attribute information estimated by the server device can be used.

  Exemplary embodiments of an attribute estimation program and an attribute information providing system according to the present invention will be described below in detail with reference to the accompanying drawings.

  In the following first embodiment, the outline and features of the attribute estimation device that executes the attribute estimation program according to the present invention, the configuration of the attribute estimation device, and the effects of the first embodiment will be described in order.

[Outline and features of attribute estimation device]
First, the outline and features of the attribute estimation apparatus according to the first embodiment will be described with reference to FIG. FIG. 1 is a system configuration diagram illustrating an overall configuration of a system including an attribute estimation apparatus according to the first embodiment. As shown in the figure, in this system, a terminal device 10 and a search server device 20 are connected to be communicable with each other via the Internet, and the search server device 20 and the attribute estimation device 30 are connected via a network. It is configured to be communicable with each other.

  In such a system, the terminal device 10 receives an input or selection of a search word from a user and transmits the search word to the search server 20, and the search server 20 is based on the search word received from the terminal device 10. Performing a WEB search and transmitting a search result (a list of URLs of WEB sites corresponding to the search term) to the terminal device 10, and the attribute estimation device 30 estimating the attribute of the user of the terminal device 10. In particular, the attribute estimation device 30 is characterized mainly in that the attribute of the user of the terminal device 10 can be accurately estimated.

  The main feature will be specifically described. The attribute estimation device 30 stores a search term input or selected by the user from the terminal device 10 and a user ID for uniquely identifying the user in association with each other. A search word log database (search word log DB) 32a is provided. Specifically, when a search term is input from the terminal device 10 to the search server 20 by the user, the search term is associated with the user ID and stored as “01, new graduate recruitment / nurse”.

  Further, the attribute estimation device 30 includes an attribute database (attribute DB) 32b that stores a predetermined attribute and a search term that is highly likely to be input or selected by a user having the attribute in association with each other. Specifically, predetermined attributes and search terms are stored in advance as “20s, new graduate recruitment / graduation ceremony / graduate school” or “female, nurse / divination”.

  In addition, the attribute estimation device 30 includes a user information database (user information DB) 32c that stores the estimated user attributes and user IDs in association with each other. Specifically, “01, 20s / female” is stored in association with the estimated user attribute and user ID.

  The terminal device 10 stores “01” as a user ID for uniquely identifying the user. Specifically, the user ID assigned by the attribute estimation device 30 to uniquely identify the user via the search server device 20 is temporarily written and stored using, for example, Cookie. is doing.

  In such a configuration, when a certain search word is input or selected in the terminal device 10 accessing the search server device 20 ((1) in FIG. 1), the search server device 20 is stored in the terminal device 10. The acquired user ID and the input search term are acquired ((2) in FIG. 1). To give a specific example, when the search term “new graduate recruitment, nurse” is input in the terminal device 10 with the user ID “01”, the search server device 20 is stored in the terminal device 10. The search term “new graduate recruitment, nurse” input with the ID “01” is acquired.

  Subsequently, the search server device 20 transmits the acquired user ID and search word to the attribute estimation device 20 ((3) in FIG. 1), and the attribute estimation device 30 receives the received user ID and search word. Are stored in the search term log DB 32a in association with each other ((4) in FIG. 1). Specifically, in the above example, the search server device 20 transmits the acquired user ID “01” and the search term “new graduate recruitment, nurse” to the attribute estimation device 20, and the attribute estimation device 30. Stores the received user ID “01” and the search term “new graduate recruitment, nurse” in the search term log DB 32a in association with each other.

  And the attribute estimation apparatus 30 estimates a user's attribute based on a match with the search word memorize | stored in search word log DB32a, and the search word memorize | stored in attribute DB32b ((5) of FIG. 1). For example, the attribute estimation unit 33a stores “new graduate recruitment, nurse” entered as a search word by the user with the user ID “01” in the user DB 32a, and “attributes, search” in the attribute DB 32b. When “20s, new graduate recruitment / graduation ceremony / graduate school” and “female, nurse / fortune telling” are stored as “word”, the search term of the attribute “20s” is “new graduate recruitment” entered by the user Since “Nurse” matches the search term of the attribute “female”, the attribute of the user ID “01” that has entered “new hiring, nurse” as the search term is set to “20s, female ".

  Thereafter, the attribute estimation device 30 stores the estimated user attribute and the user ID in association with each other in the user information DB 32c ((6) in FIG. 1). Specifically, in the above example, (01, 20s / female) is stored in the user information DB 32c in association with the estimated user attribute and user ID.

  As described above, according to the first embodiment, as in the main feature described above, a search word that is less likely to contain noise as a result of reflecting the user's intention, for example, a word that represents the same thing but has different expressions between men and women. As a result of being able to estimate attributes (age group, gender, residential area, etc.) from words used by a specific age group (for example, nurses and nurses) (for example, recruiting new graduates) It is possible to estimate well with the system.

[Configuration of Attribute Estimation Device (Example 1)]
Next, the configuration of the attribute estimation apparatus 30 shown in FIG. 1 will be described with reference to FIGS. 2 is a block diagram illustrating a configuration of the attribute estimation device 30, FIG. 3 is a diagram illustrating an example of a search term log DB, and FIG. 4 is a diagram illustrating an example of an attribute DB. FIG. 5 is a diagram illustrating an example of the user information DB. As shown in FIG. 2, the attribute estimation device 30 includes a communication control I / F unit 31, a storage unit 32, and a control unit 33.

  Among these, the communication control IF unit 14 is a means for controlling communication related to various information exchanged with the search server device 32. Specifically, the search term input from the terminal device 10 and the terminal device 10 The user ID assigned to the user is received from the search server device 20.

  The storage unit 32 is a storage unit (storage unit) that stores data and programs necessary for various processes performed by the control unit 33. In particular, the storage unit 32 is closely related to the present invention. A DB 32b and a user information DB 32c are provided. The search term log DB 32a corresponds to the “search term storage means” described in the claims. Similarly, the attribute DB 32b corresponds to the “attribute storage means”, and the user information DB 32c includes “user attribute”. Corresponds to "memory means".

  Among these, the search term log DB 32a is a means for storing a search term input or selected by the user from the terminal device 10 in association with a user ID for uniquely identifying the user. As illustrated in FIG. 3, the user ID and the search word are associated with each other and “01, new graduate recruitment / nurse” and the like are stored.

  The attribute DB 32b is a means for storing a predetermined attribute and a search term that is highly likely to be input or selected by a user having the attribute from the terminal device, specifically, as illustrated in FIG. As described above, a predetermined attribute and a search word are associated with each other, and “20s, new graduate recruitment / graduation ceremony / graduate school”, “female, nurse / divination”, and the like are stored.

  Further, the user information DB 32c is a means for storing a user attribute and a user ID estimated by the attribute estimation unit 33a described later in association with each other. Specifically, as illustrated in FIG. The person ID and the estimated attribute are associated with each other and stored as “01, female / 20s”.

  The control unit 33 has an internal memory for storing a control program such as an OS (Operating System), a program defining various processing procedures, and required data, and executes various processes using these. In particular, as closely related to the present invention, an attribute estimation unit 33a is provided. The attribute estimation unit 33a corresponds to “attribute estimation means” recited in the claims.

  The attribute estimation unit 33a is a means for estimating a user attribute based on a match between a search word stored in the search word log DB 32a and a search word stored in the attribute DB 32b. Specifically, “new graduate recruitment, nurse” entered as a search word by the user with the user ID “01” is stored in the user DB 32a (see FIG. 3), and “attribute, search word” is stored in the attribute DB 32b. When “20s, new graduate recruitment / graduation ceremony / graduate school” and “female, nurse, fortune telling” are stored (see FIG. 4), the search term “attributed to“ 20s ”is“ new graduate recruitment ”entered by the user. And “Nurse” matches the search term of the attribute “female”, the attribute estimation unit 33a has the user ID “01” that has entered “New graduate recruitment, nurse” as the search term. The user attribute is estimated as “20s, female”.

  The attribute estimation unit 33a stores the user attribute and the user ID estimated as described above in the user information DB 32c in association with each other. Note that the attribute estimation processing by the attribute estimation unit 33a is attribute estimation from the administrator via the timing at which the attribute estimation request is received from the search server device 20 or the operation unit (not shown in FIG. 2) of the attribute estimation device 30. It is executed at an arbitrary timing such as a timing when a request is input or every predetermined time interval.

[Processing by attribute estimation device]
Next, attribute estimation processing by the attribute estimation apparatus will be described with reference to FIG. FIG. 6 is a flowchart showing the flow of attribute estimation processing.

  As illustrated in FIG. 6, when the attribute estimation device 30 receives an attribute estimation request from the search server device 20 (Yes in step S601), the attribute estimation device 30 obtains a user ID and a search word corresponding to the user ID from the search word log DB 32a. In addition to reading, the combination of the search word and the attribute is read from the attribute DB 32b (step S602).

  Subsequently, the attribute estimation device 30 estimates the attribute of the user based on the match between the search word read from the search word log DB 32a and the search word read from the attribute DB 32b (step S603). For example, “new graduate recruitment and nurse” entered by the user with the user ID “01” as a search word is stored in the user DB 32a (see FIG. 3), and “attributes and search words” are stored in the attribute DB 32b as “20s. , “New graduate recruitment / graduation ceremony / graduate school” and “Women, nurses / fortune telling” are stored (see FIG. 4), the “new graduate recruitment” entered by the user matches the search term of the attribute “20s”. In addition, since “nurse” matches the search term of the attribute “female”, the attribute estimation unit 33a determines the attribute of the user with the user ID “01” who entered “new graduate recruitment, nurse” as the search term. Are estimated to be “female in their 20s”.

  Then, the attribute estimation apparatus 30 stores the estimated user attribute (for example, twenties, female) and the user ID (for example, 01) in association with each other in the user information DB 32c (step S604).

[Effect of Example 1]
As described above, according to the first embodiment, the search term input or selected by the user from the terminal device 10 is stored in association with the user ID for uniquely identifying the user, A search term that is likely to be input or selected from the terminal device 10 by a user having the attribute is stored in association with the search term stored in the search term log DB 32a and the search term stored in the attribute DB 32b. Since the attribute of the user is estimated based on the match and the estimated attribute of the user and the user ID are stored in association with each other, a search term that is less likely to contain noise as a result of reflecting the user's intention, For example, attributes (age group, gender, residential area) such as words that represent the same thing but have different expressions between men and women (for example, nurses and nurses) and words used by a specific age group (for example, new graduate recruitment) Etc.) It may result, it is possible to estimate better system attributes of the user of the terminal device.

  By the way, in the first embodiment, a case has been described in which a user's attribute is estimated using a search word stored in advance in the attribute DB 32b. However, the present invention is not limited to this and is stored in the search word log DB 32a. The search term used for attribute estimation may be used recursively in subsequent attribute estimation. In the first embodiment, a case where a single attribute is stored in the attribute DB 32b has been described. However, the present invention is not limited to this, and conflicting attributes are stored in the attribute DB 32b. Also good.

  Therefore, in the second embodiment, the search terms used for attribute estimation are used recursively in the subsequent attribute estimation using FIGS. 7 to 10, and the search terms are associated with each conflicting attribute. Will be described. In the following, the configuration of the attribute estimation apparatus, the flow of attribute estimation processing, and the effects of the second embodiment will be described in order. 7 is a flowchart illustrating the flow of attribute estimation processing according to the second embodiment, FIG. 8 is a diagram illustrating an example of a search term log DB according to the second embodiment, and FIG. 9 is related to the second embodiment. FIG. 10 is a diagram illustrating an example of a user information DB according to the second embodiment.

[Configuration of Attribute Estimation Device (Example 2)]
First, the configuration of the attribute estimation apparatus 30 according to the second embodiment will be described. Since the configuration is basically the same as that of the first embodiment, an attribute DB 32b and an attribute estimation unit that are different from the first embodiment are described below. Only 33a will be described. In the second embodiment, as shown in FIG. 8, the search term log DB 32a stores a plurality of search terms in association with each of a plurality of users. Also, as shown in FIG. In the information DB 323c, a plurality of attributes are associated with each other and stored.

  The attribute DB 32b associates and stores a search term for each conflicting attribute. Specifically, as shown in FIG. 9, search terms are stored for each conflicting attribute such as “male” and “female”, “20s” and “30s”.

  Further, the attribute DB 32b newly stores a search word and an attribute used for estimation by attribute estimation in association with each other. As a specific example, a predetermined user group is estimated as “male” by the attribute estimation process, and the majority of users included in the predetermined user group used the search term “PC”. In such a case, the attribute “male” and the search term “sea” are newly stored in the attribute DB 32b. The search word stored in the search word log DB 32a and used for attribute estimation is used recursively in subsequent attribute estimation.

  The attribute estimation unit 33a is a means for estimating user attributes. Specifically, the attribute of the user and the attribute of the search term are newly supplemented to the attribute DB 32b, and the attribute of the user is estimated recursively. Details of the processing by the attribute estimation unit 33a will be described later with reference to FIG.

[Processing by Attribute Estimation Device (Example 2)]
Next, attribute estimation processing by the attribute estimation device 30 will be described with reference to FIG. As shown in FIG. 7, when the attribute estimation device 30 receives an attribute estimation request from the outside (Yes in step S701), the attribute estimation device 30 can search the attribute DB 32b with a search word group A1 that can be estimated with high accuracy. The search term group B1 that can be estimated with high accuracy that the attribute is not α is read out and set in the work area (step S702). Specifically, from the attribute DB 32b shown in FIG. 9, it is estimated with high accuracy that the search term group A1 can be estimated with high accuracy that the attribute is “male” and the attribute is “female”. Each possible search term group B1 is read out and set in the work area.

  Subsequently, the attribute estimation device 30 sets i = 1 (step S703), and searches from the search word log DB 32a for a user group Ci searching for Ai and a user searching for Bi that conflicts with the attributes of Ai. Each group Dj is extracted (step S704). As a specific example, from the search word log DB 32a shown in FIG. 8, a user group searching for “police officer” is extracted as Ci, and also searching for “female police officer”. The user group is extracted as Dj. In this example, the user group Ci is estimated as the attribute “male”, and the user group Di is estimated as the attribute “female”.

  Then, the attribute estimation device 30 searches the search term log DB 32a for a search term group in which K% or more of the user group Ci performs a common search and L% or more of the user group Dj does not perform a search. Extract as group Ai + 1 (step S705). To give a specific example, K% or more (for example, 80% or more) of the user group Ci searching for “police officer” performs a common search, and also searches for “police officer”. The search term “sea” is used as a search term group Ai + 1 as a search term in which L% or more (for example, 80% or more) of the user group Dj who has not performed a search (ie, “female police officer”) has not performed a search. (See FIG. 8). In this example, the attribute “male” and the search term “sea” are newly stored in the attribute DB 32b.

  Subsequently, the attribute estimation device 30 searches the search term log 32a for a search term group in which K% or more of the user group Dj performs a search in common and L% or more of the user group Ci does not perform a search. Extracted as the word group Bi + 1 (step S706). To give a specific example, K% or more (for example, 80% or more) of users Dj who are searching for “female police officers” commonly search, and “female police officers” are also searching. The search term “fortune” is set as the search term group Bi + 1 as a search term in which L% or more (for example, 80% or more) of the user group Dj not (that is, searching for “police officer”) is not searching. Extract (see FIG. 8). In this example, the attribute “female” and the search term “fortune telling” are newly stored in the attribute DB 32b.

  Then, the attribute estimation apparatus 30 determines whether i is equal to or greater than j (step S707). If i is not equal to or greater than j (No in step S707), the process returns to step S704, and the user groups Ci and Dj are extracted again. Then, the processes in steps S705 to S707 are repeated. That is, in the above-described example, the user group Ci and the user group Di are newly extracted based on the search terms “sea” and “fortune telling” extracted in the process of step S704, A new search term is extracted from the search term log 32a by the processing of S705 and S706.

  On the other hand, if i is greater than or equal to j (Yes at Step S707), the attribute estimation device 30 sets the attribute of the estimated user group Ci to “α” for each process of Step S704 that is repeatedly executed, The estimation result that the attribute of the user group Di is “not α” is stored in the user information DB 32c (step S708). Specifically, as shown in FIG. 10, “02, male / 20's” “10, female” is stored.

[Effect of Example 2]
As described above, according to the second embodiment, the attribute DB 32a newly stores the search word and the attribute used for estimation in association with each other, and the attribute estimation unit 33a newly searches the search word log DB 32a. User attributes are estimated using words and attributes. By adding the search terms and attributes used for user attribute estimation, the coverage of the predetermined attributes is improved. It is possible to further improve the accuracy of the results.

  Further, according to the second embodiment, the attribute DB 32a stores a search term in association with each conflicting attribute, and the attribute estimation unit 33a is stored in the search term and attribute DB 32a stored in the search term log DB 32a. Since the user attribute is estimated based on the match with the search term for each conflicting attribute, the result of the estimation using conflicting attributes of the search term (for example, male, female, etc.) It is possible to obtain an estimation result with higher accuracy than in the case of estimation using only attributes.

  In the first and second embodiments described above, the case where only the attribute is acquired as the estimation result has been described. However, the present invention is not limited to this, and the estimation accuracy indicating the accuracy of the estimated attribute is calculated. You may do it. Therefore, in the following, as the third embodiment, a case where the estimation accuracy is calculated will be described. Note that this estimation accuracy calculation processing is executed following the processing in steps S705 and S706, for example, in the flowchart shown in FIG.

  For example, as illustrated in FIG. 11, the attribute estimation device 30 sets “an” as a probability that a user whose attribute is A inputs a search word n, and sets a probability that a user whose attribute is not A as input a search word n. bn ", the probability that a user whose attribute is A inputs the search word m is" am ", and the probability that a user who does not have the attribute A inputs the search word m is" bm ", and then the attribute is A A probability “anm” in which the user inputs both the search word n and the search word m, and a probability “bnm” in which a user whose attribute is not A input the search word n and the search word m are calculated, respectively, and the search word n The probability that the attribute of the user who has input both the search term m is A is calculated as “anm / (anm + bnm)”. That is, when the user whose attribute “A” is estimated inputs both the search word n and the search word m, the accuracy of the estimated attribute “A” is calculated as described above. Become.

  Then, the attribute estimation device 30 stores the estimation accuracy calculated as described above in association with the user ID in the user information DB 32c together with the attribute. Specifically, as shown in FIG. 12, in association with the user ID, the estimated gender, gender estimation accuracy, estimated age group, age group estimation accuracy, estimated residence area, The estimated accuracy of the residential area is stored in the user information DB 32c (see (1) in the figure). In addition, although the example which estimated sex, an age group, and a residence area was shown here as an attribute, these are only examples, and as shown in the same figure, other attributes (occupation, preference, etc.) are estimated. In addition, the user information may be stored in an expansion table.

  Thus, according to the third embodiment, as a result of calculating the estimation accuracy in association with the attribute and user ID, the accuracy of the estimation result of each user ID or attribute can be known, and the accuracy of the estimation result can be determined. It can be a guideline for further improvement.

  Incidentally, in the above-described embodiment, the attribute DB and search word storage processing for the attribute DB 32a is not specified, but the search word input or selected by the user from the terminal device 10 whose attribute is known in advance is stored in the attribute DB 32b. You may make it store.

  Therefore, as a fourth embodiment, a case where a user attribute (user attribute) is estimated using a search term from the terminal device 10 whose attribute is known in advance will be described with reference to FIG. FIG. 13 is an overall configuration diagram of a system including an attribute estimation device using a mobile phone network.

  As shown in FIG. 13, the terminal device 130 and the search server 131 are connected to be communicable with each other via the Internet, the search server 131 and the attribute estimation device 135 are connected, and the mobile phone 133 The search server 134 is connected to be communicable with each other via a mobile phone network, and the search server 134 and the attribute estimation device 135 are connected. Note that the mobile phone 133 is, for example, a personal computer, a workstation, a home game machine, an Internet TV, a PDA, or a mobile communication terminal such as a mobile phone or a PHS.

  Here, since the mobile phone 133 requires personal information to make a contract, the attributes of the user (user) are known in advance. When the user accesses the search server 134 from the mobile phone 133 and inputs a search term, the search server 134 assigns a user ID uniquely to the user of the mobile phone 133 and associates the user ID with the search term. Store in the log DB.

  Then, the attribute estimation device 135 stores the search term input or selected by the user from the mobile phone 133 whose attribute is known in advance in association with the predetermined attribute in the attribute DB, and is stored in the search term log DB. The attribute of the user (user) is estimated based on a match between the search term and the search term input or selected by the user whose attribute is stored in advance in the attribute DB. The search term log DB corresponds to “search term storage means” described in the claims, and similarly, the attribute DB corresponds to “attribute storage means”.

  For example, when the user of the mobile phone 133 who is “20 years old, male” inputs the search term “employment”, the attribute estimation device 135 uses “employment” as the search terms for “20 years old” and “male” in the attribute DB. ”And the attribute of the user (user) is estimated based on the match between the search word stored in the search word log DB and the search word input by the user (user). Then, as an estimation result, the attributes of the user (user) such as “20 years old, male, student” are estimated in more detail.

  As described above, according to the fourth embodiment, the attribute DB 32b stores the search term input or selected by the user from the mobile phone 133 whose attribute is known in advance in association with the predetermined attribute, and the attribute estimation device. 135 is based on a match between a search word stored in the search word log DB and a search word input or selected by a user (user) whose attribute stored in the attribute DB is known in advance. Since the attribute is estimated, for example, when a search word from a terminal device whose attribute is known is used as a result of estimating the attribute using a search word from a terminal device such as a mobile phone, the attribute is estimated in advance. It is possible to obtain an estimation result with higher accuracy than.

  By the way, in the above-described embodiment, the case where the user attribute (user attribute) is estimated using only the search term from the terminal device 10, the present invention is not necessarily limited to this, and the terminal device 10 is not limited thereto. The user attribute (user attribute) may be estimated by further using the URL information and the search term browsed from or entered and / or the IP address of the terminal device 10.

  Specifically, as shown in FIG. 14, the search term log DB 32a stores the date and time when the user (user) executed the search or the date and time when the URL was selected from the search result, and the access source IP address of the user (user). , A unique identifier of the user (user), a search term input by the user (user), and URL information when the user (user) selects a URL from the search result, “Data, IP, User_ID, Keyword, URL "is stored. For example, “2006/1/1, 192.168.1.1, 0001, entrance examination, http://www.xxx.yyy.com” is stored. Then, the attribute estimation apparatus 30 according to the fifth embodiment estimates the user's attribute using not only the search word but also the above-described URL information, search time, and the IP address of the terminal device 10.

  As described above, according to the fifth embodiment, the attribute estimation unit further estimates the user's attribute by using the URL information browsed from the terminal device by the user. As a result of using the search term, it is possible to obtain an estimation result with higher accuracy than in the case of estimation using only the search term.

  Further, according to the fifth embodiment, the attribute estimation unit estimates the user's attribute by further using the time when the search word is input or selected by the user from the terminal device and / or the IP address of the terminal device. As a result of further estimation using the time when the search word is input or selected and / or the IP address of the terminal device, it is possible to obtain an estimation result with higher accuracy than in the case of estimation using only the search word. .

  By the way, the present invention may provide the estimated attribute information to another device or system. Therefore, in the sixth embodiment, a system using attribute information estimated by the attribute estimation device will be described with reference to FIGS. 15 and 16. FIG. 15 is a diagram illustrating an overall configuration of a service example (information provision) using attribute information according to the sixth embodiment. FIG. 16 is a diagram illustrating a service example (advertisement provision) using attribute information according to the sixth embodiment. It is a figure which shows the whole structure.

[Service example of attribute estimation processing (information provision)]
As shown in FIG. 15, a terminal device 150 for inputting a search word by a user (user), a WWW server 151 for providing a WEB page according to the input search word, and a user (user) of the terminal device 150 ) Is provided so as to be able to communicate with each other via the Internet.

  The information providing server 152 includes a user information DB that stores user attributes and user IDs in association with each other. Further, the terminal device 150 holds the uniquely assigned user ID in the cookie file.

  Under such a configuration, when the user accesses the WWW server 151 from the terminal device 150 and transmits a browsing request for the WEB page ((1) in FIG. 15), the WWW server 151 determines the WWW server ID and the final value. A transfer URL including the URL to be accessed and the transaction code is returned to the terminal device 150 ((2) in FIG. 15).

  Subsequently, the terminal device 150 adds the user ID in the cookie file to the received transfer URL and accesses the information providing server 152 ((3) in FIG. 15). Then, the information providing server 152 holds the received transfer URL and instructs the transfer to the URL to be finally accessed ((4) in FIG. 15).

  Thereafter, the terminal device 150 accesses the instructed URL ((5) in FIG. 15). Then, the accessed web server (not shown) notifies the information providing server of its WWW server ID and transaction code ((6) in FIG. 15).

  Thereafter, the information providing server 152 identifies the user ID from the received WWW server ID and transaction code, extracts the attribute from the user information DB, and transmits the extracted attribute to the web server ((7) in FIG. 15). . Then, the web server that has received the attribute displays a WEB page such as content suitable for the attribute on the terminal device 150 ((8) in FIG. 15).

[Service example of attribute estimation processing (advertisement provision)]
As shown in FIG. 16, a terminal device 160 that browses a WEB page by a user (user), a WWW server 161 that provides a WEB page according to an input search word, and a user (user) of the terminal device 160 ) Is provided so as to be able to communicate with each other via the Internet.

  The advertisement providing server 162 includes a user information DB that stores user attributes and user IDs in association with each other. In addition, the terminal device 160 holds a uniquely assigned user ID in the cookie file.

  As shown in FIG. 16, when the user accesses the WWW server 151 from the terminal device 160 and transmits a web page browsing request ((1) in FIG. 16), the WWW server 161 accesses the advertisement providing server 162. Is transmitted to the terminal device 160 ((2) in FIG. 16).

  Subsequently, the terminal device 160 that has received the response transmits the user ID of the cookie file to the advertisement providing server ((3) in FIG. 16). Then, the advertisement providing server 162 extracts an attribute corresponding to the received user ID from the user information DB, and transmits an advertisement suitable for the extracted attribute to the terminal device 160 ((4) in FIG. 16).

[Effects of Example 6]
As described above, according to the sixth embodiment, the information stored in the user information DB is provided to other terminal devices, so that, for example, attribute information can be shared among a plurality of other sites such as a search engine. It is not necessary to estimate the attribute at each site, and it is possible to suppress capital investment when performing a service that requires attribute information.

  Further, according to the sixth embodiment, the information providing server 152 associates and stores a search term input or selected by the user from the terminal device 150 and a user ID for uniquely identifying the user, A predetermined attribute and a search term that is likely to be input or selected by the user having the attribute from the terminal device 150 are stored in association with each other and stored in the search term storage unit and the attribute storage unit. User attribute is estimated based on the match with the search term, the estimated user attribute and the user ID are stored in association with each other, and a request relating to the user attribute is received from the WWW server 151 In this case, the estimated user attribute is transmitted to the WWW server 151, and the WWW server 151 transmits a request related to the user attribute of the terminal device 150 to the information providing server 152, and the user attribute is Since the predetermined information is transmitted to the terminal device 150 based on the received user attributes received from the service server 152, even when the WWW server 151 and the information providing server 152 belong to different domains (for example, search service When the attribute estimation service is performed by different companies), the user ID is not known to the WWW server 151, and the service using the attribute information estimated by the information providing server 152 can be performed.

  Further, according to the sixth embodiment, the advertisement providing server 162 stores a search term input or selected by the user from the terminal device 160 in association with a user ID for uniquely identifying the user, A search term stored in the search term storage unit and a search term stored in the attribute storage unit are stored in association with a predetermined attribute and a search term that is likely to be input or selected by a user having the attribute. The user attribute is estimated based on the match, the estimated user attribute and the user ID are stored in association with each other, and the terminal device 160 accessing the WWW server 161 changes the user attribute to the user attribute. When such a request is received, predetermined information corresponding to the estimated user attribute is transmitted to the terminal device 160, and the WWW server 161 accepts access from the terminal device 160 and sends the access response. Therefore, when the WWW server 161 and the advertisement providing server 162 belong to the same domain (for example, when the search service and the advertisement service are performed by the same company), the cost can be reduced when the system is constructed without constructing a complicated system configuration. And the attribute information estimated by the advertisement providing server 162 can be used.

  Although the embodiments of the present invention have been described so far, the present invention may be implemented in various different forms other than the embodiments described above. Accordingly, another embodiment included in the present invention will be described below as a seventh embodiment.

(1) System configuration, etc. Further, all or part of the processes described as being automatically performed among the processes (for example, FIG. 6 and FIG. 7) by the attribute estimation apparatus described in the first to sixth embodiments are manually performed. Alternatively, all or part of the processing described as being performed manually can be automatically performed by a known method. In addition, the processing procedure, control procedure, specific name, and information including various data and parameters shown in the above-described document and drawings can be arbitrarily changed unless otherwise specified.

  Each component of the attribute estimation apparatus shown in FIG. 2 is functionally conceptual and does not necessarily need to be physically configured as illustrated. That is, the specific form of the distribution / integration of the attribute estimation device (for example, the search word log DB 32a as an external device) is not limited to that shown in the figure, and all or a part thereof can be used for various loads and usage situations. Accordingly, it can be configured to be functionally or physically distributed and integrated in arbitrary units. Further, all or any part of each processing function performed in each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware by wired logic.

(2) Program By the way, the various processes described in the above embodiments can be realized by executing a program prepared in advance on a computer system such as a personal computer or a workstation. Therefore, in the following, an example of a computer system that executes a program having the same function as in the above embodiment will be described.

[Computer system that executes attribute estimation program]
FIG. 17 is a diagram illustrating a computer system 170 that executes an attribute estimation program. As shown in the figure, the computer system 170 includes a RAM 171, an HDD 172, a ROM 173, and a CPU 174. Here, the ROM 173 stores in advance an attribute estimation program 173a, as shown in FIG.

  Then, the CPU 174 reads out and executes the program 173a, thereby forming an attribute estimation process 174a as shown in FIG. Note that the attribute estimation process 174a corresponds to the attribute estimation unit 33a illustrated in FIG.

  The HDD 172 also stores a search word storage table 172a for storing a search word input or selected by the user from the terminal device and a user ID for uniquely identifying the user, and predetermined attributes. An attribute estimation table 172b that stores a search term that is highly likely to be input or selected from the terminal device by a user having the attribute, a user attribute estimated by the attribute estimation program 173a, and a user ID And a user attribute table 172c for storing them in association with each other. The search term storage table 172a corresponds to the search term log DB 32a shown in FIG. 2, similarly, the attribute estimation table 172b corresponds to the attribute DB 32b, and the user attribute table 172c corresponds to the user information DB 32c. To do.

  The attribute estimation program 173a is not necessarily stored in the ROM 173. For example, a flexible disk (FD), a CD-ROM, an MO disk, a DVD disk, a magneto-optical disk, which is inserted into the computer system 170, In addition to “portable physical media” such as IC cards, “fixed physical media” such as hard disk drives (HDDs) provided inside and outside the computer system 60, public lines, the Internet, LAN, WAN, etc. The program may be stored in “another computer system” connected to the computer system 170, and the computer system 170 may read out and execute the program from these.

(Supplementary note 1) An attribute estimation program for causing a computer to execute a method for estimating a user attribute of a terminal device,
A search word storage procedure for storing a search word input or selected from the terminal device by the user and a user ID for uniquely identifying the user in a search word storage unit;
An attribute storage procedure for associating a predetermined attribute with a search term that is likely to be input or selected by the user having the attribute from the terminal device in the attribute storage unit;
An attribute estimation procedure for estimating an attribute of the user based on a match between a search word stored in the search word storage unit and a search word stored in the attribute storage unit;
A user attribute storage procedure for storing the user attribute estimated by the attribute estimation procedure and the user ID in association with each other in a user attribute storage unit;
Attribute estimation program that causes a computer to execute.

(Additional remark 2) The said attribute storage procedure matches the search term and attribute used for estimation by the said attribute estimation procedure, and newly stores them in the said attribute storage means,
The attribute estimation program according to appendix 1, wherein the attribute estimation procedure causes a computer to estimate the attribute of the user using a search word and an attribute newly stored in the search word storage unit.

(Additional remark 3) It further has an estimation accuracy calculation means for calculating an estimation accuracy indicating the accuracy of the attribute estimated by the attribute estimation procedure,
The attribute estimation program according to appendix 1 or 2, wherein the user attribute storage procedure causes the computer to execute further storing the estimation accuracy calculated by the estimation accuracy calculation unit in association with the user attribute storage unit.

(Supplementary Note 4) The attribute storing procedure stores a search term input or selected by a user from the terminal device whose attribute is known in advance in association with the predetermined attribute in the attribute storage unit,
The attribute estimation procedure is based on a match between a search word stored in the search word storage unit and a search word input or selected by a user whose attribute stored in the attribute storage unit is known in advance. The attribute estimation program according to any one of supplementary notes 1 to 3, which causes a computer to estimate the attribute of the user.

(Additional remark 5) The said attribute storage procedure matches each search word for every conflicting attribute, and stores it in the said attribute storage means,
The attribute estimation procedure estimates the attribute of the user based on a match between a search word stored in the search word storage unit and a search word for each conflicting attribute stored in the attribute storage unit. The attribute estimation program according to any one of appendices 1 to 3, which causes a computer to execute

(Additional remark 6) The said attribute estimation procedure is any one of additional remark 1-5 which makes a computer perform estimating the said user's attribute further using the URL information browsed from the said terminal device by the said user. Attribute estimation program described in 1.

(Additional remark 7) The said attribute estimation procedure estimates the said user's attribute further using the time when the said search word was input or selected from the said terminal device by the user, and / or the IP address of the said terminal device. The attribute estimation program according to any one of appendices 1 to 6, which causes a computer to execute the above.

(Additional remark 8) The attribute estimation program as described in any one of additional remarks 1-7 which makes a computer further perform the estimated information provision procedure which provides the information stored in the said user attribute memory | storage means to another terminal device.

(Supplementary note 9) It is configured by connecting a terminal device used by a user, a web server device that provides information to the terminal device of the user, and an information providing server device that estimates the attributes of the user to a network. Attribute information providing system,
The information providing server device includes:
Search term storage means for storing a search term input or selected by the user from the terminal device and a user ID for uniquely identifying the user;
Attribute storage means for storing a predetermined attribute and a search term that is highly likely to be input or selected by the user having the attribute from the terminal device;
Attribute estimation means for estimating an attribute of the user based on a match between a search word stored in the search word storage means and a search word stored in the attribute storage means;
User attribute storage means for storing the attribute of the user estimated by the attribute estimation means and the user ID in association with each other;
An attribute transmission unit that transmits the attribute of the user estimated by the attribute estimation unit to the web server device when a request related to the attribute of the user is received from the web server device;
The web server device
An attribute receiving means for transmitting a request related to the attribute of the user of the terminal device to the information providing server device, and receiving the attribute of the user from the information providing server device;
Information transmitting means for transmitting predetermined information to the terminal device based on the attribute of the user received by the attribute receiving means;
An attribute information providing system characterized by comprising:

(Supplementary Note 10) A terminal device used by a user, a web server device that provides information to the user's terminal device, and an information providing server device that estimates the user's attributes are connected to a network. Attribute information providing system,
The information providing server device includes:
Search term storage means for storing a search term input or selected from the terminal device by the user from the terminal device and a user ID for uniquely identifying the user;
Attribute storage means for storing a predetermined attribute and a search term that is highly likely to be input or selected by a user having the attribute;
Attribute estimation means for estimating an attribute of the user based on a match between a search word stored in the search word storage means and a search word stored in the attribute storage means;
User attribute storage means for storing the attribute of the user estimated by the attribute estimation means and the user ID in association with each other;
When a request related to the user attribute is received from the terminal device accessing the web server device, the terminal device receives predetermined information corresponding to the user attribute estimated by the attribute estimation means. Information transmitting means for transmitting to
The web server device
Access response means for accepting access from the terminal device and performing the access response;
An attribute information providing system characterized by comprising:

(Additional remark 11) It is an attribute estimation apparatus which estimates the attribute of the user of a terminal device,
A search word storage means that associates a search word input or selected from the terminal device by the user with a user ID for uniquely identifying the user;
Attribute storage means for storing a predetermined attribute and a search term that is highly likely to be input or selected by the user having the attribute from the terminal device;
Attribute estimation means for estimating an attribute of the user based on a match between a search word stored in the search word storage means and a search word stored in the attribute storage means;
User attribute storage means for storing the attribute of the user estimated by the attribute estimation means and the user ID in association with each other;
An attribute estimation device comprising:

(Supplementary note 12) An attribute estimation method suitable for estimating an attribute of a user of a terminal device,
A search term storage step of associating a search term input or selected by the user from the terminal device with a user ID for uniquely identifying the user in the search term storage means;
An attribute storage step of associating a predetermined attribute with a search term that is highly likely to be input or selected by the user having the attribute from the terminal device;
An attribute estimation step for estimating the attribute of the user based on a match between the search word stored in the search word storage means and the search word stored in the attribute storage means;
A user attribute storage step of storing the user attribute estimated by the attribute estimation step and the user ID in association with each other in a user attribute storage unit;
An attribute estimation method characterized by comprising:

  As described above, the attribute estimation program, the attribute estimation device, the attribute estimation method, and the attribute estimation system according to the present invention are useful for estimating the attributes of the user of the terminal device. Suitable for estimating attributes accurately.

1 is an overall configuration diagram of a system including an attribute estimation device according to Embodiment 1. FIG. It is a block diagram which shows the structure of the attribute estimation apparatus which concerns on Example 1. FIG. It is a figure which shows the structural example of the information memorize | stored in search word log DB which concerns on Example 1. FIG. It is a figure which shows the structural example of the information memorize | stored in attribute DB which concerns on Example 1. FIG. It is a figure which shows the structural example of the information memorize | stored in user information DB which concerns on Example 1. FIG. 6 is a flowchart illustrating a flow of attribute estimation processing according to the first embodiment. 12 is a flowchart illustrating a flow of attribute estimation processing according to the second embodiment. It is a figure which shows the structural example of the information memorize | stored in search word log DB which concerns on Example 2. FIG. It is a figure which shows the structural example of the information memorize | stored in attribute DB which concerns on Example 2. FIG. It is a figure which shows the structural example of the information memorize | stored in user information DB which concerns on Example 2. FIG. It is a figure which shows the example of the calculation method of estimation accuracy. It is a figure which shows the structural example of the information memorize | stored in user information DB which concerns on Example 3. FIG. It is a whole block diagram of the system containing the attribute estimation apparatus using the mobile telephone network which concerns on Example 4. FIG. It is a figure which shows the structural example of the information memorize | stored in search word log DB which concerns on Example 5. FIG. It is a figure which shows the whole structure of the example of a service (information provision) using the attribute information which concerns on Example 6. FIG. It is a figure which shows the whole structure of the example of a service (advertisement provision) using the attribute information which concerns on Example 6. FIG. It is a figure which shows the example of the computer system which performs an attribute estimation program.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Terminal device 20 Search server apparatus 30 Attribute estimation apparatus 31 Communication control I / F part 32 Storage part 32a Search term log DB
32b Attribute DB
32c User information DB
130 terminal device 131 search server 132 WWW server 133 mobile phone 134 search server 135 attribute estimation device 150 terminal device 151 WWW server 152 information providing server 160 terminal device 161 WWW server 162 information providing server 170 computer system 171 RAM
172 HDD
172a Search term storage table 172b Attribute estimation table 172c User attribute table 173 ROM
173a Attribute estimation program 174 CPU
174a Attribute estimation process

Claims (5)

An attribute estimation program for causing a computer to execute a method for estimating a user attribute of a terminal device,
A search word storage procedure for storing a search word input or selected from the terminal device by the user and a user ID for uniquely identifying the user in a search word storage unit;
An attribute storage procedure for associating a predetermined attribute with a search term that is likely to be input or selected by the user having the attribute from the terminal device in the attribute storage unit;
An attribute estimation procedure for estimating an attribute of the user based on a match between a search word stored in the search word storage unit and a search word stored in the attribute storage unit;
A user attribute storage procedure for storing the user attribute estimated by the attribute estimation procedure and the user ID in association with each other in a user attribute storage unit;
A search word associated with a specific attribute stored in the attribute storage means is selected, and a search is performed with a first user group that has performed a search with the selected search word and a search word that is in conflict with the specific attribute. The second user group is extracted from the search word storage means, the first user group searches for a predetermined value or more, and the second user group searches for a predetermined value or more. A storage procedure for specifying a search term that has not been performed from the search term storage unit, and storing the specified search term and the specific attribute in association with each other in the attribute storage unit
An attribute estimation program for causing a computer to execute. The attribute storage procedure newly stores the search word stored in the search word storage unit in the attribute storage unit in association with the attribute estimated by the attribute estimation procedure,
The attribute estimation program according to claim 1, wherein the attribute estimation procedure estimates the attribute of the user by using a search word and an attribute newly stored in the search word storage unit. Causing the computer to further execute an estimation accuracy calculation procedure for calculating an estimation accuracy indicating the accuracy of the attribute estimated by the attribute estimation procedure;
The attribute estimation program according to claim 1 or 2, wherein the user attribute storing procedure further stores the estimated accuracy calculated by the estimated accuracy calculating procedure in association with the user attribute storage means. Attribute information provision configured by connecting to a network a terminal device used by a user, a web server device that provides information to the terminal device of the user, and an information providing server device that estimates the user's attributes A system,
The information providing server device includes:
Search term storage means for storing a search term input or selected by the user from the terminal device and a user ID for uniquely identifying the user;
Attribute storage means for storing a predetermined attribute and a search term that is highly likely to be input or selected by the user having the attribute from the terminal device;
Attribute estimation means for estimating an attribute of the user based on a match between a search word stored in the search word storage means and a search word stored in the attribute storage means;
User attribute storage means for storing the attribute of the user estimated by the attribute estimation means and the user ID in association with each other;
An attribute transmission unit that transmits the attribute of the user estimated by the attribute estimation unit to the web server device when a request related to the attribute of the user is received from the web server device;
A search word associated with a specific attribute stored in the attribute storage means is selected, and a search is performed with a first user group that has performed a search with the selected search word and a search word that is in conflict with the specific attribute. The second user group is extracted from the search word storage means, the first user group searches for a predetermined value or more, and the second user group searches for a predetermined value or more. Storage means for identifying a search term that has not been performed from the search term storage means, and storing the specified search term and the specific attribute in association with each other in the attribute storage means,
The web server device
An attribute receiving means for transmitting a request related to the attribute of the user of the terminal device to the information providing server device, and receiving the attribute of the user from the information providing server device;
Information transmitting means for transmitting predetermined information to the terminal device based on the attribute of the user received by the attribute receiving means;
An attribute information providing system characterized by comprising: Attribute information provision configured by connecting to a network a terminal device used by a user, a web server device that provides information to the terminal device of the user, and an information providing server device that estimates the user's attributes A system,
The information providing server device includes:
Search term storage means for storing a search term input or selected by the user from the terminal device and a user ID for uniquely identifying the user;
Attribute storage means for storing a predetermined attribute and a search term that is highly likely to be input or selected by the user having the attribute from the terminal device ;
Attribute estimation means for estimating an attribute of the user based on a match between a search word stored in the search word storage means and a search word stored in the attribute storage means;
User attribute storage means for storing the attribute of the user estimated by the attribute estimation means and the user ID in association with each other;
When a request related to the user attribute is received from the terminal device accessing the web server device, the terminal device receives predetermined information corresponding to the user attribute estimated by the attribute estimation means. Information transmitting means for transmitting to
A search word associated with a specific attribute stored in the attribute storage means is selected, and a search is performed with a first user group that has performed a search with the selected search word and a search word that is in conflict with the specific attribute. The second user group is extracted from the search word storage means, the first user group searches for a predetermined value or more, and the second user group searches for a predetermined value or more. Storage means for identifying a search term that has not been performed from the search term storage means, and storing the specified search term and the specific attribute in association with each other in the attribute storage means,
The web server device
Access response means for accepting access from the terminal device and performing the access response;
An attribute information providing system characterized by comprising:

Images