JP2019144906A - Information processing apparatus, information processing method, and program - Google Patents

Abstract

To automatically identify a web page corresponding to a life event.SOLUTION: The information processing apparatus includes in the event-related page specifying unit: a timing specifying unit that specifies an occurrence timing of predetermined life event of each of a plurality of users; a first browsing history acquisition unit that acquires web-page browsing history information of one or more first users for which the occurrence timing is specified; a browsing page identifying unit that, for each of the one or more first users, based on the browsing history information, identifies a web page browsed in a first period that is a period before the occurrence timing, and a web page browsed in a second period that is a period after the occurrence timing; and a related-page identifying unit that identifies a web page related to the life event from among a plurality of web pages based on a change in a browsing status of each of the one or more first users in each of the first period and the second period.SELECTED DRAWING: Figure 3

Description

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

  It is known that when a life event, which is an event related to a user's life or life, occurs, access to information related to the life event increases rapidly, and purchases of products or services related to the life event increase rapidly. . According to the occurrence of a user's life event, by distributing an advertisement related to a product or service related to the life event, a highly efficient advertisement can be realized, and therefore the user who has generated the life event is specified. .

  For example, Patent Literature 1 discloses a system that estimates that a life event has occurred when a web page viewed by a user matches a web page corresponding to a life event.

JP 2017-117351 A

  However, the system disclosed in Patent Document 1 discloses how the web page corresponding to the life event is specified, although the web page corresponding to the life event is stored in the database in advance. Not. Although it is conceivable to manually specify a web page corresponding to a life event, manual identification is complicated. Therefore, it is required to accurately specify a web page corresponding to a life event.

  Accordingly, the present invention has been made in view of these points, and an object thereof is to provide an information processing apparatus, an information processing method, and a program capable of automatically specifying a web page corresponding to a life event. .

  In the information processing apparatus according to the first aspect of the present invention, the generation timing of the life event is specified by the timing specifying unit that specifies the generation timing of each predetermined life event of the plurality of users, and the timing specifying unit. For each of the one or more first users, a first browsing history acquisition unit that indicates browsing history of one or more first users' web pages and acquires browsing history information including an address and browsing time of the web page, Based on the browsing history information, with the identified occurrence timing as a reference, a web page browsed in a first period that is a period before the occurrence timing and a second period that is a period after the occurrence timing The browsing page specifying unit for specifying the web page browsed in the site, and before the browsing page specifying unit specified A web page related to the life event from the plurality of web pages based on a change in the browsing status of the web page in each of the first period and the second period of each of the one or more first users; A related page specifying unit for specifying an event related page.

  The related page specifying unit includes a plurality of web page browsing states in each of the first period and the second period of the one or more first users specified by the browsing page specifying unit. The number of the first users who browsed each of the web pages during the first period and the number of the first users who browsed during the second period are calculated, and the second number is calculated for each of the plurality of web pages. A ratio of the number of the first users in the first period to the number of the first users corresponding to the period is calculated, and among the plurality of web pages, a web page whose ratio exceeds a first threshold value, You may identify with the pre-event related page which is the said event related page with high possibility of being browsed before generation | occurrence | production of the said life event.

  The related page specifying unit includes a plurality of web page browsing states in each of the first period and the second period of the one or more first users specified by the browsing page specifying unit. Calculating the number of browsing of the one or more first users in the first period of each of the web pages and the number of browsing of the one or more first users in the second period, and each of the plurality of web pages The ratio of the number of browsing in the first period to the number of browsing in the second period is calculated, and among the plurality of web pages, a web page in which the ratio exceeds a second threshold is You may identify with the pre-event related page which is the said event related page with high possibility of being browsed before generation | occurrence | production.

  The browsing page specifying unit specifies a web page browsed in at least one of the plurality of first periods, and the related page specifying unit includes a plurality of the first periods for each of the plurality of web pages. The number of the first users corresponding to each of the plurality of first periods is calculated with respect to the number of the first users corresponding to the second period. A web page in which at least one of the plurality of calculated ratios exceeds the first threshold among the plurality of web pages may be identified as the pre-event related page.

  The first browsing history acquisition unit acquires the browsing history information of one or more second users whose generation timing is not specified by the timing specifying unit, and the browsing page specifying unit is the one or more second items. For each user, a web page browsed in the first period based on at least one of the one or more occurrence timings specified for the one or more first users based on the browsing history information The related page specifying unit determines a ratio of the number of the first users corresponding to the first period to the number of the first users corresponding to the second period for each of the plurality of web pages. The number of the first users corresponding to the first period with respect to the number of the second users corresponding to the first period A second ratio that is a ratio is calculated, and among the plurality of web pages, a web page in which the first ratio exceeds a first threshold and the second ratio exceeds a third threshold is set as the life event. A related pre-event related page may be specified.

  The information processing apparatus includes: a second browsing history acquisition unit that acquires the browsing history information of one or more third users to be identified by the life event by the timing specifying unit; and the one or more third users For each of the plurality of third periods, which are periods before the current time, based on the browsing history information for each of the number of browsing one or more unrelated pages different from the pre-event related pages A third ratio that is a ratio of the number of browsing related pages is calculated, and based on the plurality of third ratio changes calculated for each of the one or more third users, from the one or more third users, You may further provide the user specific part which specifies the user with high possibility that the said life event will generate | occur | produce.

  For each of the one or more third users, the user specifying unit calculates an increase rate of the third ratio based on two adjacent third periods in order from the past third period in the plurality of third periods. You may calculate and specify the 3rd user whose increase rate of the calculated several 3rd ratio exceeds a 4th threshold value as a user with high possibility that the said life event will generate | occur | produce.

  The first browsing history acquisition unit acquires the browsing history information of one or more second users whose generation timing has not been specified by the timing specifying unit, and the information processing device is configured to acquire the browsing of the first user. Among the history information, the browsing history information before the generation timing is used as the teacher data of the user who has generated the life event, and the browsing history information of the second user is the user's data for which the life event has not occurred. A generation unit that generates a model for classifying whether or not a user corresponding to the browsing history information is a user who is highly likely to generate a life event with respect to the input of the browsing history information as teacher data; The second browsing history collection for acquiring the browsing history information of one or more third users to be specified by the timing specifying unit as the life event. And the third user's browsing history information is input to the model, and the third user is classified into the model as to whether or not the user is highly likely to generate a life event. You may further provide the user specific part which specifies the user with high possibility that the said life event will generate | occur | produce from the above 3rd user.

  The related page specifying unit calculates, for each of the plurality of web pages, a ratio of the number of the first users corresponding to the second period to the number of the first users corresponding to the first period; Of the plurality of web pages, a web page in which the ratio exceeds the fifth threshold may be identified as a post-event related page that is the event-related page that is likely to be browsed after the life event occurs. .

  The browsing page specifying unit specifies a web page browsed in at least one of the plurality of second periods, and the related page specifying unit includes a plurality of second periods for each of the plurality of web pages. The number of the first users corresponding to each of the plurality of second periods is calculated with respect to the number of the first users corresponding to the first period. Then, among the plurality of web pages, a web page in which at least one of the calculated plurality of ratios exceeds the fifth threshold may be specified as the post-event related page.

  The information processing apparatus is different from the post-event related pages in each of the plurality of third periods that are periods before the current time, based on the browsing history information, for each of the one or more third users. A fourth ratio that is a ratio of the number of browsing of the related page after the event to the number of browsing of the one or more non-related pages is calculated, and the plurality of the fourth ratios calculated for each of the one or more third users are calculated. You may further provide the user specific part which specifies the user who the said life event was complete | finished from the said one or more 3rd users based on a change.

  The related page specifying unit calculates, for each of the plurality of web pages, a ratio of the number of first users corresponding to the second period to the number of first users corresponding to the first period, Among the plurality of web pages, a web page in which the ratio exceeds the fifth threshold is identified as a post-event related page that is the event-related page that is highly likely to be browsed after the occurrence of the life event, and the information processing The apparatus may further include an advertisement distribution unit that distributes an advertisement corresponding to the post-event related page that is likely to be viewed after the life event to a user who is highly likely to generate the life event.

  The information processing apparatus classifies the pre-event related page and the post-event related page into categories to which each page belongs, and associates a period during which a page corresponding to the category is likely to be browsed with the category. The advertisement distribution unit further includes a classification unit, based on the browsing history information of a user who is likely to generate the life event, the category of the web page that the user is currently browsing relatively frequently. The advertisement corresponding to the category identified and related to the period later than the period associated with the identified category may be distributed.

  The information processing method according to the second aspect of the present invention includes a step of identifying a predetermined life event occurrence timing of each of a plurality of users executed by a computer, and one or more of which the occurrence timing of the life event is identified A browsing history of the first user's web page, obtaining browsing history information including the web page address and browsing time, and for each of the one or more first users based on the browsing history information A web page browsed in a first period that is a period before the occurrence timing, and a web page browsed in a second period that is a period after the occurrence timing, with the occurrence timing specified as a reference Identifying the first period of each of the identified one or more first users; Based on the change in the viewing conditions of a web page in each of the serial second period, from the plurality of web pages, and a step of specifying an event related pages to web pages associated with the life events.

  In the program according to the third aspect of the present invention, a computer is used to specify the generation timing of a predetermined life event for each of a plurality of users, and the generation timing of the life event is specified by the timing specification unit. A browsing history acquisition unit that displays browsing history of one or more first users' web pages and acquires browsing history information including an address and browsing time of the web page, and the browsing history for each of the one or more first users. Based on the information, on the basis of the identified occurrence timing, a web page browsed in a first period that is a period before the occurrence timing and a second page that is a period after the occurrence timing are browsed A browsing page specifying unit for specifying a web page and the browsing page specifying unit The plurality of web pages are related to the life event based on a change in browsing status of the web pages in the first period and the second period of each of the one or more first users. It functions as a related page specifying unit that specifies an event related page as a web page.

  According to the present invention, there is an effect that a web page corresponding to a life event can be automatically specified.

It is a figure which shows the outline | summary of the information processing apparatus which concerns on 1st Embodiment. It is a figure which shows the structure of the information processing apparatus which concerns on 1st Embodiment. It is a figure which shows the structure of an event related page specific | specification part. It is a figure which shows the example of the 1st ratio corresponding to each of a some web page. It is a figure which shows the example of the 2nd ratio corresponding to each of a some web page. It is a figure which shows the structure of an event generation user specific part. It is a figure which shows the structure of an advertisement delivery part. It is a flowchart which shows the flow of a process in the information processing apparatus when specifying a pre-event related page. It is a flowchart which shows the flow of a process in information processing apparatus when specifying a user with high possibility that a life event will generate | occur | produce. It is a flowchart which shows the flow of a process in information processing apparatus when delivering an advertisement to a user with high possibility that a life event will generate | occur | produce. It is a figure which shows the structure of the event generation user specific | specification part which concerns on 2nd Embodiment.

[Outline of Information Processing Apparatus 1]
FIG. 1 is a diagram illustrating an overview of an information processing apparatus 1 according to the first embodiment. The information processing apparatus 1 identifies a web page related to a life event, identifies a user who is likely to generate a life event based on a change in the browsing status of the user's web page, and advertises the user to the user. It is a computer that delivers. Here, the life event is an event related to a user's life or life, and examples thereof include moving, marriage, employment, car purchase, and the like.

  The information processing apparatus 1 acquires event occurrence date information indicating an occurrence date as an event occurrence timing through a questionnaire or the like from a plurality of users U1 ((1) in FIG. 1). Further, the information processing apparatus 1 acquires browsing history information indicating browsing histories of a plurality of web pages from the first user terminal 2 used by each of one or more users U1 ((2) in FIG. 1). The information processing apparatus 1 identifies an event-related page that is a web page related to a life event based on changes in the browsing status of the web pages before and after the event occurrence date of the plurality of users U1 ((3) in FIG. 1). ). By doing so, the information processing apparatus 1 can automatically specify the event related page.

The information processing apparatus 1 acquires browsing history information from the second user terminal 3 used by each of one or more users U2 to be identified as life events ((4) in FIG. 1). The information processing apparatus 1 identifies a user who is highly likely to generate a life event based on a change in the browsing status of the event-related page ((5) in FIG. 1), and a user who is highly likely to generate a life event. The advertisement is distributed to (6 in FIG. 1). By doing in this way, information processor 1 can raise an advertising effect.
Hereinafter, the configuration of the information processing apparatus 1 will be described.

[Configuration Example of Information Processing Apparatus 1]
FIG. 2 is a diagram illustrating a configuration of the information processing apparatus 1 according to the first embodiment. The information processing apparatus 1 includes a storage unit 11 and a control unit 12.

  The storage unit 11 is, for example, a ROM (Read Only Memory) and a RAM (Random Access Memory). The storage unit 11 stores various programs for causing the information processing apparatus 1 to function. For example, the storage unit 11 stores a program that causes the control unit 12 of the information processing device 1 to function as an event-related page specifying unit 13, an event occurrence user specifying unit 14, and an advertisement distribution unit 15 described later.

  The control unit 12 is, for example, a CPU (Central Processing Unit). The control unit 12 controls functions related to the information processing apparatus 1 by executing various programs stored in the storage unit 11. The control unit 12 functions as an event related page specifying unit 13, an event occurrence user specifying unit 14, and an advertisement distribution unit 15 by executing a program stored in the storage unit 11.

[Identify event-related pages]
The event related page specifying unit 13 specifies an event related page based on the browsing histories of the plurality of first users U1. The event-related page specifying unit 13 can be viewed as a web page related to a life event, a pre-event related page that is an event-related page that is highly likely to be viewed before the occurrence of the life event, and a life event. A post-event related page that is a highly event related page is identified.

  FIG. 3 is a diagram showing a configuration of the event related page specifying unit 13. As shown in FIG. 3, the event related page specifying unit 13 includes a timing specifying unit 131, a first browsing history acquisition unit 132, a browsing page specifying unit 133, and a related page specifying unit 134.

  The timing specifying unit 131 specifies the occurrence date of each predetermined life event for each of the plurality of first users. For example, the timing specifying unit 131 acquires questionnaire result information indicating a result of a questionnaire regarding a life event, which is performed for each of the plurality of users U1. The questionnaire result information includes a user ID for identifying the user U1, type information indicating each type of the plurality of life events, and occurrence date information indicating the occurrence dates of the plurality of life events.

  Here, in the user U1, there are a user who has generated a predetermined life event and a user who has not generated the predetermined life event. A user U1 in which a predetermined life event has occurred is referred to as a first user, and a user U1 in which a predetermined life event has not occurred is referred to as a second user. The timing specifying unit 131 specifies the occurrence date of each predetermined life event of the plurality of first users based on the acquired questionnaire result information. In the following description, a predetermined life event is simply referred to as a life event. Moreover, in this embodiment, although the timing specific | specification part 131 specified the generation | occurrence | production date of each predetermined life event of several 1st users, it does not restrict to this. The timing specifying unit 131 may specify the generation week and the generation time of the predetermined life event as the generation timing of the predetermined life event.

  The first browsing history acquisition unit 132 acquires browsing history information indicating browsing histories of one or more first users' web pages whose life event occurrence dates have been identified by the timing identification unit 131, and the life event occurrence date. Browsing history information of one or more second users for which is not specified is acquired. Specifically, the 1st browsing history acquisition part 132 acquires browsing history information corresponding to each of a plurality of users U1 who answered a questionnaire about a life event. The browsing history information includes a user ID for identifying the user U1, an address of the browsed web page, and browsing time information indicating the browsing time. Here, the address of the web page is assumed to be a URL, for example.

  The browsing page identification unit 133 is browsed for each of one or more first users in a first period that is a period before the occurrence date based on the occurrence date of the identified life event based on the browsing history information. And the web page browsed in the second period, which is the period after the date of occurrence. By doing in this way, the time axis of the browsing date of each of a some 1st user's web page can be unified on the basis of the occurrence date of a life event.

Here, a plurality of first periods are provided, and the browsing page specifying unit 133 specifies a web page browsed in at least one of the plurality of first periods for each of the one or more first users. The first period is, for example, 24 hours. In the following description, each of the plurality of first periods is expressed as T ₋₁ , T ₋₂ ,..., T _−n from the order close to the life event occurrence date. Further, the plurality of first periods are used for specifying the pre-event related page.

In addition, a plurality of second periods are provided, and the browse page specifying unit 133 specifies a web page browsed in at least one of the plurality of second periods for each of the one or more first users. The second period is, for example, 24 hours. In the following description, each of the plurality of second periods is expressed as T ₁ , T ₂ ,... T _{n in} order from the order close to the life event occurrence date. Further, the plurality of second periods are used for specifying the post-event related page.

  The browsing page specifying unit 133 is based on at least one of the one or more occurrence dates specified for the one or more first users based on the browsing history information for each of the one or more second users. The web pages browsed in a plurality of first periods before the date of occurrence are specified.

  For example, when the occurrence date of the predetermined life event specified for the first user X is January 30, 2018, the browsing page specifying unit 133 sets the browsing history information for each of the one or more second users. The web pages browsed in a plurality of first periods before the date of occurrence are specified based on the date of occurrence. In addition, for example, when the occurrence date of the predetermined life event specified for the first user Y is February 10, 2018, the browsing page specifying unit 133 browses each of the one or more second users. Based on the history information, the web page browsed in a plurality of first periods before the date of occurrence is specified based on the date of occurrence.

  The related page specifying unit 134 includes a plurality of web pages based on changes in the browsing status of the web pages in each of the first period and the second period of the one or more first users specified by the browsing page specifying unit 133. An event-related page to be a web page related to a life event is specified from the page. The related page specifying unit 134, as a web page related to a life event, is a pre-event related page that is an event related page related to the occurrence of a life event, and a post-event related page that is an event related page related to the end of the life event. And specify. First, a method for specifying a pre-event related page will be described.

[Specify related page before event]
First, the related page specifying unit 134 sets each of the plurality of web pages in the first period based on the browsing status of the web pages in the first period of each of the one or more first users specified by the browsing page specifying unit 133. The number of first users who browsed is calculated. The related page specifying unit 134 calculates the number of first users who browsed each of the plurality of first periods for each of the plurality of web pages.

Moreover, the related page specific | specification part 134 calculates the number of the 1st users who browsed each of several web pages in the 2nd period based on the browsing condition of the web page in the 2nd period. Here, the relevant page identification unit 134, for each of the plurality of web pages, to calculate the number of the first user who has viewed the nearest second period T ₁ to the date of life events.

Then, the relevant page identification unit 134, for each of the plurality of web pages, to the number of the first user corresponding to the second period T _1, the number ratio of the first user corresponding to each of the plurality of first period A certain first ratio is calculated.

  FIG. 4 is a diagram illustrating a first ratio corresponding to each of a plurality of web pages. As shown in FIG. 4, a web page having a relatively high first ratio is likely to be viewed by a first user who generates a life event before the life event occurs, and may be a pre-event related page. high.

  Further, the related page specifying unit 134 sets each of the plurality of web pages to the first period based on the browsing status of the web pages in the first period of each of the one or more second users specified by the browsing page specifying unit 133. The number of second users who browsed is calculated. Then, the related page specifying unit 134 calculates a second ratio that is a ratio of the number of first users corresponding to the first period to the number of second users corresponding to the first period. The related page specifying unit 134 calculates the second ratio for each of the plurality of first periods.

  FIG. 5 is a diagram illustrating a second ratio corresponding to each of a plurality of web pages. For example, a web page having a relatively high second ratio is more likely to be viewed by a first user who has experienced a life event, and is likely to be an event-related page.

  The related page specifying unit 134 specifies the pre-event related page from the plurality of web pages based on the plurality of first ratios and the plurality of second ratios calculated for each of the plurality of web pages.

  For example, the related page specifying unit 134 includes, among the plurality of web pages, at least one of the plurality of calculated first ratios exceeding the first threshold and at least one of the plurality of second ratios being the third. Web pages that exceed the threshold are identified as pre-event related pages that are likely to be viewed before the occurrence of a life event. Note that the related page specifying unit 134 includes, among the plurality of web pages, at least one of the calculated first ratios exceeding the threshold th1 (first threshold), or the calculated plurality of web pages. A web page in which at least one of the second ratios exceeds the threshold th2 (third threshold) may be specified as the pre-event related page.

Here, the threshold th1 is 1.5, for example. The related page specifying unit 134 may determine the threshold th2 based on the number of first users and the number of second users. For example, when the number of first users is M and the number of second users is N, the threshold th2 is expressed by the following equation (1).
th2 = 2 * M / N (1)

  In this embodiment, the ratio of the number of first users in the first period to the number of first users corresponding to the second period is the first ratio, and the pre-event related page is specified based on the first ratio. However, it is not limited to this.

  The related page specifying unit 134 includes a plurality of web pages based on the browsing status of the web pages in each of the first period and the second period of the one or more first users specified by the browsing page specifying unit 133. You may calculate the browsing number of the 1 or more 1st user in each 1st period, and the browsing number of the 1 or more 1st user in the 2nd period. And the related page specific | specification part 134 may calculate the ratio of the browsing number in a 1st period with respect to the browsing number in a 2nd period about each of a some web page. And the related page specific | specification part 134 may specify the web page in which the said ratio exceeds threshold value th3 (2nd threshold value) among several web pages with the pre-event related page. By doing in this way, information processor 1 can specify a pre-event related page similarly to specifying a pre-event related page using the number of users of the 1st user.

[Identification of related page after event]
Next, a method for specifying a post-event related page will be described.
First, the related page specifying unit 134 sets each of the plurality of web pages in the first period based on the browsing status of the web pages in the first period of each of the one or more first users specified by the browsing page specifying unit 133. The number of first users who browsed is calculated. Here, the related page specifying unit 134 calculates, for each of the plurality of web pages, the number of first users who have viewed the web page in the first period T ₋₁ that is closest to the occurrence date of the life event.

  Moreover, the related page specific | specification part 134 calculates the number of the 1st users who browsed each of several web pages in the 2nd period based on the browsing condition of the web page in the 2nd period. The related page specifying unit 134 calculates the number of first users who browsed each of the plurality of second periods for each of the plurality of web pages.

And the related page specific | specification part 134 is ratio of the number of the 1st users corresponding to each of several 2nd period with respect to the number of 1st users corresponding to 1st period T- ₁ about each of several web pages. Is calculated. Then, the related page specifying unit 134 specifies, as a post-event related page, a web page in which at least one of the plurality of calculated ratios exceeds the threshold th4 (fifth threshold) among the plurality of web pages.

[Identify users who are likely to have life events]
The event occurrence user specifying unit 14 specifies a user who is highly likely to generate a life event from one or more third users, based on browsing history information of one or more third users to be specified as life events. . FIG. 6 is a diagram illustrating a configuration of the event occurrence user specifying unit 14. As shown in FIG. 6, the event occurrence user identification unit 14 includes a second browsing history acquisition unit 141 and a user identification unit 142.

  The 2nd browsing history acquisition part 141 acquires browsing history information which shows a browsing history of a web page of one or more 3rd users U2 made into a specific subject of a life event.

For each of one or more third users, the user specifying unit 142, based on the browsing history information acquired by the second browsing history acquisition unit 141, in each of a plurality of third periods that are periods before the current time, A third ratio P3, which is a ratio of the number of browsing of the pre-event related page to the number of browsing of one or more non-related pages different from the pre-event related page, is calculated. In the following description, each of the plurality of third periods is expressed as T3 ₋₁ , T3 _{−2 ,.}

  And the user specific | specification part 142 selects a user with high possibility that a life event will generate | occur | produce from one or more 3rd users based on the change of several 3rd ratio calculated about each of one or more 3rd users. Identify.

Specifically, for each of one or more third users, the user specifying unit 142 sequentially enters two adjacent third periods T3 _-i (where 0 ≦ 3) in order from the past third period in the plurality of third periods. A change rate (increase rate) of the third ratio P3 in i ≦ n−1) is calculated. When the third ratio in the third period T3- _i is P3- _i and the third ratio in the third period T3- _{(i + 1)} is P3- _{(i + 1)} , the rate of change is P3- _i / P3- _{(i + 1)} . Represented.

  When the change rate of the third ratio P3 indicates an increase in the third ratio, the user specifying unit 142 specifies a user corresponding to the third ratio as a user who is highly likely to generate a life event. . For example, the user specifying unit 142 specifies a third user whose rate of change of the third ratio P3 exceeds the threshold th5 (fourth threshold) as a user who is likely to generate a life event.

In addition, the user identification part 142 is the 3rd ratio P3 of several 3rd period before the said 3rd period in order from the past 3rd period in several 3rd period in order of each of one or more 3rd users. The variance σ _p3 and the average value μ _p3 may be calculated. And user specific part 142 may specify a user with high possibility that a life event will occur based on the distribution σ _p3 and average value μ _p3 . For example, user identification unit 142, when the third ratio P3 _-i in the third period T3 _-i satisfies the following equation (2), possibly life event occurs is identified as the high user.

[Identify users who are most likely to have finished their life events]
The event occurrence user specifying unit 14 may have ended the life event from one or more third users based on the browsing history information of the third user, in the same manner as specifying a user who is likely to generate a life event. Identify highly likely users.

  Specifically, for each of one or more third users, the user specifying unit 142 is based on the browsing history information acquired by the second browsing history acquisition unit 141, and includes a plurality of third times that are periods before the current time. A fourth ratio P4, which is a ratio of the number of browsing related pages after the event to the number of browsing one or more non-related pages different from the related page after the event in each period, is calculated, and each of the one or more third users is calculated. Based on the changes in the plurality of fourth ratios P4 calculated for the user, a user who is highly likely to have finished the life event is identified from one or more third users.

Specifically, for each of one or more third users, the user specifying unit 142 sequentially enters two adjacent third periods T3 _-i (where 0 ≦ 3) in order from the past third period in the plurality of third periods. A change rate (increase rate) of the fourth ratio P4 in i ≦ n−1) is calculated. When the fourth ratio in the third period T3- _i is P4- _i and the fourth ratio in the third period T3- _{(i + 1)} is P4- _{(i + 1)} , the rate of change is P4- _i / P4- _{(i + 1)} . Represented.

  When the change rate of the fourth ratio P4 indicates an increase in the fourth ratio, the user specifying unit 142 specifies a user corresponding to the fourth ratio as a user who is highly likely to have finished the life event. For example, the user specifying unit 142 specifies a third user whose rate of change of the fourth ratio P4 exceeds the threshold th7 as a user who is highly likely to have finished the life event.

  Here, when the change rate of the third ratio P3 indicates a decrease and the change rate of the fourth ratio P4 indicates an increase of the fourth ratio, the user specifying unit 142 indicates that the third ratio and the third ratio You may identify the user corresponding to 4 ratios as a user with high possibility that the life event was complete | finished.

In addition, for each of the one or more third users, the user specifying unit 142 sequentially starts from the past third period in the plurality of third periods, and the fourth ratio P4 of the plurality of third periods before the third period. The variance σ _p4 and the average value μ _p4 may be calculated. And the user specific | specification part 142 may specify the user with high possibility that the life event was complete _| finished based on the said dispersion _| distribution (sigma) _p4 and average value (mu) _p4 . For example, user identification unit 142, when the fourth rate P4 _-i in the third period T3 _-i satisfies the equation (3) below, to identify the possibility of life event is completed is high user.

[Distribution of advertisements to users who are likely to have life events]
The advertisement distribution unit 15 is an advertisement corresponding to a post-event related page that is likely to be viewed after the life event to the third user who is identified by the event occurrence user identification unit 14 and is likely to experience the life event. To deliver. FIG. 7 is a diagram illustrating a configuration of the advertisement distribution unit 15. As shown in FIG. 7, the advertisement distribution unit 15 includes a classification unit 151 and a distribution unit 152.

  The classification unit 151 classifies the pre-event related page and the post-event related page into categories to which the respective pages belong. Specifically, the classification unit 151 identifies the category to which each of the pre-event related page and the post-event related page belongs. For example, the classification unit 151 specifies a category by acquiring a category assigned to the pre-event related page and the post-event related page by a search engine or the like. Note that the classification unit 151 may identify a category to which these pages belong by analyzing words included in the pre-event related page and the post-event related page.

  The classification unit 151 identifies a period during which a page corresponding to each of a plurality of categories is likely to be browsed, and associates the identified period with the category. For example, the classification unit 151 identifies a period during which the pre-event related page and the post-event related page classified into the category are most viewed based on the browsing history information of the first user acquired by the first browsing history acquisition unit 132. To do. The classification unit 151 specifies a period in which the pre-event related page and the post-event related page are most viewed based on the life event occurrence date. Then, the classification unit 151 causes the storage unit 11 to store category information that associates the identified period with the category.

  Based on the browsing history information of the user who is likely to generate a life event and the classification result by the classification unit 151, the distribution unit 152 selects the category of the web page that the user is currently browsing relatively frequently. Identify. For example, the distribution unit 152 identifies a category corresponding to the web page that is currently browsed most.

  The storage unit 11 stores a category and advertisement information indicating an advertisement corresponding to the category in association with each other. The distribution unit 152 refers to the category information stored in the storage unit 11 and identifies a category associated with a period later than the period associated with the identified category. And the delivery part 152 delivers the advertisement linked | related with the said category to the user with high possibility that a life event will generate | occur | produce.

  For example, the user identification unit 142 stores the user ID of a user who is likely to generate a life event in the storage unit 11. The distribution unit 152 functions as a DSP (Demand Side Platform) in advertisement distribution, and identifies a user ID of the user when the user visits a web page. The distribution unit 152 refers to the storage unit 11 and determines whether or not the identified user ID is stored as a user ID of a user who is likely to generate a life event.

  When the distribution unit 152 determines that the user is a user who is highly likely to generate a life event, the distribution unit 152 is later than the period associated with the identified category in the advertising space of the web page visited by the user. Serve ads associated with the category associated with the date range.

[Flow of processing in information processing apparatus 1]
Subsequently, an example of a processing flow in the information processing apparatus 1 will be described. First, the flow of processing in the information processing apparatus 1 when specifying a pre-event related page will be described. FIG. 8 is a flowchart showing the flow of processing in the information processing apparatus 1 when specifying a pre-event related page.

First, the timing specifying unit 131 specifies the occurrence date of each predetermined life event for each of the plurality of first users (S1).
Subsequently, the first browsing history acquisition unit 132 acquires browsing history information of the first user and the second user (S2).

Subsequently, the browsing page specifying unit 133 and the related page specifying unit 134 calculate the number of first users who browsed each of the plurality of web pages for each of the plurality of first periods and the second period (S3).
Subsequently, the related page specifying unit 134 calculates the first ratio of each of the plurality of web pages (S4).

Subsequently, the browsing page specifying unit 133 and the related page specifying unit 134 calculate the number of second users who browsed each of the plurality of web pages for each of the plurality of first periods (S5).
Subsequently, the related page specifying unit 134 calculates the second ratio of each of the plurality of web pages (S6).
Subsequently, the related page specifying unit 134 specifies the pre-event related page from the plurality of web pages based on the calculated first ratio and second ratio (S7).

  Next, the flow of processing when the information processing apparatus 1 specifies a user who is likely to generate a life event will be described. FIG. 9 is a flowchart showing a flow of processing in the information processing apparatus 1 when a user who is likely to generate a life event is specified. This flowchart is performed by each of a plurality of third users. In addition, in this flowchart, for simplification of description, the flow of processing when attention is paid to one previous event-related page will be described.

First, the second browsing history acquisition unit 141 acquires browsing history information indicating the browsing history of the web page of the third user that is the target of the life event (S11).
Subsequently, the user specifying unit 142 refers to the acquired browsing history information and corresponds to the number of browsing one or more unrelated pages different from the pre-event related pages in each of a plurality of third periods before the current time. Then, a third ratio that is a ratio of the number of browsing of the pre-event related page is calculated (S12).

Subsequently, the user specifying unit 142 sets the value of the variable i indicating the position of the third period to n in order to calculate the change rate of the third ratio in order from the past third period in the plurality of third periods. (S13).
Subsequently, the user specifying unit 142 determines whether or not the variable i is larger than 1 (S14). If the variable i is greater than 1, the user identification unit 142 moves the process to S15, and if the variable i is 1 or less, the user identification unit 142 ends this flowchart.
Subsequently, the user specifying unit 142 subtracts 1 from the value of the variable i (S15).

Subsequently, the user specifying unit 142 calculates the rate of change of the third ratio P3- _i in the third period T3- _i (S16).
Subsequently, the user specifying unit 142 determines whether or not the calculated change rate of the third ratio P3- _i is larger than the threshold th5. The user specifying unit 142 moves the process to S18 when the calculated change rate of the third ratio P3- _i is larger than the threshold th5, and moves the process to S14 when it is equal to or less than the threshold th5.

Subsequently, the user specifying unit 142 calculates the variance σ _p3 and the average value μ _p3 of the third period prior to the third period T3- _i (S18).
Subsequently, the user specifying unit 142 determines whether the third ratio P3- _i in the third period T3- _i satisfies the above-described formula (2) (S19). If the user specifying unit 142 determines that the third ratio P3- _i satisfies Expression (2), the process proceeds to S20. If the user specifying unit 142 determines that expression (2) is not satisfied, the process proceeds to S14.

  In S20, the user specifying unit 142 specifies the third user who has acquired the browsing history information in S11 as a user who is likely to generate a life event.

  Subsequently, a flow of processing when the information processing apparatus 1 distributes an advertisement to a user who is highly likely to generate a life event will be described. FIG. 10 is a flowchart illustrating a flow of processing in the information processing apparatus 1 when an advertisement is distributed to a user who is likely to generate a life event. It is assumed that category information in which a category and a period corresponding to the category are associated is stored in the storage unit 11 by the classification unit 151 before the start of the flowchart.

  First, the distribution unit 152 identifies a category corresponding to a user who is likely to generate a life event (S31). Specifically, the distribution unit 152 specifies the category of the web page that the user is browsing a relatively large amount at the present time based on the browsing history information of the user who is likely to generate a life event.

Subsequently, the distribution unit 152 refers to the category information stored in the storage unit 11 and identifies the period associated with the category identified in S31 (S32).
Subsequently, the distribution unit 152 refers to the category information stored in the storage unit 11 and identifies a category associated with a period later than the identified period (S33).
Subsequently, the distribution unit 152 refers to the advertisement information stored in the storage unit 11 and distributes the advertisement associated with the category specified in S33 to the user (S34).

[Effect in the first embodiment]
As described above, the information processing device 1 according to the first embodiment specifies the date of occurrence specified based on the browsing history information indicating the browsing history of the web page of one or more first users for which the date of occurrence of the life event is specified. The web page browsed in the first period that is a period before the date of occurrence and the web page browsed in the second period that is the period after the date of occurrence are specified and Based on the change of the browsing situation of the web page in each of the first period and the second period of the first user, an event related page that is a web page related to a life event is selected from a plurality of web pages. Identify.

  In this way, the information processing apparatus 1 automatically displays the event-related page based on the browsing history information indicating the browsing history of the web page of one or more first users whose life event occurrence dates are specified. Can be acquired. In addition, the information processing apparatus 1 reduces the influence of individual user attributes on the identification of a web page as the number of first users increases, and a general user browses in relation to a life event. It is possible to accurately identify event-related pages with high possibility.

Second Embodiment
[Use machine learning to identify users who are likely to have life events]
Next, the second embodiment will be described. The information processing apparatus 1 according to the second embodiment is different from the information processing apparatus 1 according to the first embodiment in that a user who is likely to generate a life event is specified using machine learning. Hereinafter, the information processing apparatus 1 according to the second embodiment will be described. In addition, description is abbreviate | omitted suitably about the same part as 1st Embodiment.

  FIG. 11 is a diagram illustrating a configuration of the event occurrence user specifying unit 14 according to the second embodiment. As shown in FIG. 11, the event occurrence user specifying unit 14 further includes a first browsing history acquisition unit 143 and a generation unit 144.

  Similarly to the first browsing history acquisition unit 132 according to the first embodiment, the first browsing history acquisition unit 143 includes one or more web pages of one or more first users whose life event occurrence dates are specified by the timing specification unit 131. The browsing history information indicating the browsing history is acquired, and the browsing history information of one or more second users whose life event occurrence dates were not specified is acquired.

  The generation unit 144 sets the browsing history information before the occurrence date of the life event among the browsing history information of the first user in which the life event has occurred as teacher data of the user in which the life event has occurred. The browsing history information of the second user who has not occurred is used as the teacher data of the user for whom no life event has occurred. Here, the generation unit 144 uses the date of occurrence of the life event of the predetermined first user as a reference among the browsing history information of the second user, and the life event is generated as the browsing history information before the date of occurrence. It is good also as teacher data of the user who is not.

  The generation unit 144 performs machine learning based on teacher data of a user who has generated a life event and teacher data of a user who has not generated a life event, and converts the browsing history information into the browsing history information. A model for classifying whether or not the corresponding user is a user who is highly likely to generate a life event is generated.

The second browsing history acquisition unit 141 acquires browsing history information indicating a browsing history from the current time to a predetermined period of time among browsing histories of one or more third user web pages to be identified as life events.
The user specifying unit 142 uses the browsing history information indicating the browsing history from the current time to a predetermined period before the third user's browsing history information acquired by the second browsing history acquisition unit 141 as the model generated by the generating unit 144. By inputting and classifying whether or not the third user is a user who is highly likely to generate a life event into the model, a user who is highly likely to generate a life event is selected from one or more third users. Identify.

[Effects of Second Embodiment]
As described above, the information processing apparatus 1 according to the second embodiment performs machine learning based on teacher data of a user who has generated a life event and teacher data of a user who has not generated a life event, and browse history information In response to the input, a model that classifies whether or not the user corresponding to the browsing history information is a user who is highly likely to generate a life event. Then, the information processing apparatus 1 causes the generated model to classify whether or not the third user is a user who is highly likely to generate a life event, so that the life event is received from one or more third users. Identify users who are likely to occur. By doing in this way, the information processing apparatus 1 can pinpoint the user with high possibility that a life event will generate | occur | produce with sufficient precision.

  As mentioned above, although this invention was demonstrated using said embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. It will be apparent to those skilled in the art that various modifications or improvements can be added to the above embodiment. In particular, the specific embodiments of the distribution / integration of the devices are not limited to those illustrated above, and all or a part thereof may be arbitrarily selected according to various additions or according to the functional load. It can be configured to be functionally or physically distributed and integrated in units.

DESCRIPTION OF SYMBOLS 1 ... Information processing apparatus, 11 ... Memory | storage part, 12 ... Control part, 13 ... Event related page specific part, 131 ... Timing specific part, 132 ... 1st browsing history acquisition part 133 ... browsing page identification unit, 134 ... related page identification unit, 14 ... event occurrence user identification unit, 141 ... second browsing history acquisition unit, 142 ... user identification unit, 143 First browsing history acquisition unit, 144 ... generating unit, 15 ... advertisement distribution unit, 151 ... classification unit, 152 ... distribution unit, 2 ... first user terminal, 3 ...・ Second user terminal

Claims (15)

A timing identifying unit that identifies the occurrence timing of each predetermined life event of a plurality of users;
A first browsing history that indicates browsing history of one or more first users' web pages for which the occurrence timing of the life event is specified by the timing specifying unit, and acquires browsing history information including the address and browsing time of the web page An acquisition unit;
For each of the one or more first users, based on the browsing history information, the web page browsed in a first period that is a period before the generation timing, based on the identified generation timing, A browsing page identification unit that identifies a web page browsed in the second period, which is a period after the occurrence timing;
From the plurality of web pages based on changes in the browsing status of the web pages in each of the first period and the second period of each of the one or more first users identified by the browsing page identification unit, A related page specifying unit for specifying an event related page as a web page related to the life event;
An information processing apparatus comprising: The related page specifying unit includes a plurality of web page browsing states in each of the first period and the second period of the one or more first users specified by the browsing page specifying unit. The number of the first users who browsed each of the web pages during the first period and the number of the first users who browsed during the second period are calculated, and the second number is calculated for each of the plurality of web pages. A ratio of the number of the first users in the first period to the number of the first users corresponding to the period is calculated, and among the plurality of web pages, a web page whose ratio exceeds a first threshold value, Identify the pre-event related page that is the event-related page that is likely to be viewed before the occurrence of the life event,
The information processing apparatus according to claim 1. The related page specifying unit includes a plurality of web page browsing states in each of the first period and the second period of the one or more first users specified by the browsing page specifying unit. Calculating the number of browsing of the one or more first users in the first period of each of the web pages and the number of browsing of the one or more first users in the second period, and each of the plurality of web pages The ratio of the number of browsing in the first period to the number of browsing in the second period is calculated, and among the plurality of web pages, a web page in which the ratio exceeds a second threshold is Identify the pre-event related page that is the event-related page that is likely to be viewed before occurrence,
The information processing apparatus according to claim 1. The browse page specifying unit specifies a web page browsed in at least one of the plurality of first periods,
The related page specifying unit calculates, for each of the plurality of web pages, the number of the first users corresponding to each of the plurality of first periods, and the number of the first users corresponding to the second period. The ratio of the number of the first users corresponding to each of the plurality of first periods is calculated, and at least one of the calculated plurality of ratios among the plurality of web pages has the first threshold value. Identify web pages that exceed the pre-event related pages,
The information processing apparatus according to claim 2. The first browsing history acquisition unit acquires the browsing history information of one or more second users whose generation timing is not specified by the timing specifying unit,
The browsing page specifying unit determines at least one of the one or more generation timings specified for the one or more first users based on the browsing history information for each of the one or more second users. As a reference, the web page browsed in the first period is specified,
The related page specifying unit is a ratio of the number of the first users corresponding to the first period to the number of the first users corresponding to the second period for each of the plurality of web pages. Calculating a ratio and calculating a second ratio that is a ratio of the number of the first users corresponding to the first period to the number of the second users corresponding to the first period; A web page in which the first ratio exceeds a first threshold and the second ratio exceeds a third threshold is identified as a pre-event related page related to the life event.
The information processing apparatus according to claim 2. A second browsing history acquisition unit that acquires the browsing history information of one or more third users to be specified by the timing specification unit of the life event;
For each of the one or more third users, one or more unrelated pages different from the pre-event related pages in each of a plurality of third periods that are periods before the current time based on the browsing history information. A third ratio, which is a ratio of the number of browsing pages related to the pre-event related page to the number of browsing, and based on a plurality of changes in the third ratio calculated for each of the one or more third users, A user specifying unit that specifies a user who is likely to generate the life event from one or more third users;
The information processing apparatus according to any one of claims 2 to 5. For each of the one or more third users, the user specifying unit calculates an increase rate of the third ratio based on two adjacent third periods in order from the past third period in the plurality of third periods. Calculating and identifying a third user whose calculated increase rate of the third ratio exceeds a fourth threshold as a user who is likely to generate the life event;
The information processing apparatus according to claim 6. The first browsing history acquisition unit acquires the browsing history information of one or more second users whose generation timing is not specified by the timing specifying unit,
Among the browsing history information of the first user, the browsing history information before the generation timing is used as teacher data of the user who has generated a life event, and the browsing history information of the second user is used as a life event. Model that classifies whether or not the user corresponding to the browsing history information is a user who is highly likely to generate a life event in response to the input of the browsing history information A generating unit for generating
A second browsing history acquisition unit that acquires the browsing history information of one or more third users to be specified by the timing specification unit of the life event;
The browsing history information of the third user is input to the model, and the third user is classified into the model as to whether or not the user is highly likely to generate a life event. A user identifying unit that identifies users who are likely to generate the life event from three users;
Further comprising
The information processing apparatus according to any one of claims 1 to 5. The related page specifying unit calculates, for each of the plurality of web pages, a ratio of the number of the first users corresponding to the second period to the number of the first users corresponding to the first period; Among the plurality of web pages, a web page in which the ratio exceeds a fifth threshold is identified as a post-event related page that is the event-related page that is likely to be browsed after the occurrence of the life event.
The information processing apparatus according to claim 2. The browsing page identification unit identifies a web page browsed in at least one of the plurality of second periods,
The related page specifying unit calculates, for each of the plurality of web pages, the number of first users corresponding to each of the plurality of second periods, and the number of the first users corresponding to the first period. The ratio of the number of the first users corresponding to each of the plurality of second periods is calculated, and among the plurality of web pages, at least one of the calculated ratios has the fifth threshold value. Identify web pages that exceed the relevant page after the event.
The information processing apparatus according to claim 9. For each of the one or more third users, based on the browsing history information, one or more unrelated pages different from the post-event related pages in each of a plurality of third periods that are periods before the current time. Calculating a fourth ratio that is a ratio of the number of browsing of the related page after the event to the number of browsing, and based on a plurality of changes in the fourth ratio calculated for each of the one or more third users, A user identifying unit that identifies a user who has completed the life event from one or more third users;
The information processing apparatus according to claim 10. The related page specifying unit calculates, for each of the plurality of web pages, a ratio of the number of first users corresponding to the second period to the number of first users corresponding to the first period, Among the plurality of web pages, a web page whose ratio exceeds the fifth threshold is identified as a post-event related page that is the event-related page that is likely to be browsed after the occurrence of the life event,
It further includes an advertisement distribution unit that distributes an advertisement corresponding to the post-event related page that is likely to be viewed after the life event to a user who is highly likely to generate the life event.
The information processing apparatus according to any one of claims 6 to 11. A classifying unit that classifies the pre-event related page and the post-event related page into a category to which each page belongs, and associates a period during which the page corresponding to the category is likely to be viewed with the category;
Based on the browsing history information of a user who is likely to generate the life event, the advertisement distribution unit identifies the category of the web page that the user is currently browsing relatively many times, and the identified category Delivering the ad corresponding to a category associated with a period later than the period associated with
The information processing apparatus according to claim 12. The computer runs,
Identifying the occurrence timing of a predetermined life event for each of a plurality of users;
Showing the browsing history of the web page of one or more first users for which the occurrence timing of the life event is specified, and obtaining browsing history information including the address and browsing time of the web page;
For each of the one or more first users, based on the browsing history information, the web page browsed in a first period that is a period before the generation timing, based on the identified generation timing, Identifying a web page browsed in a second period that is a period after the occurrence timing;
The plurality of web pages are related to the life event based on a change in browsing status of the web pages in the first period and the second period of each of the identified one or more first users. Identifying event-related pages as web pages to be
An information processing method comprising: Computer
A timing specifying unit for specifying the occurrence timing of each predetermined life event of a plurality of users;
A browsing history acquisition unit that displays browsing history of one or more first users' web pages for which the occurrence timing of the life event is specified by the timing specifying unit, and acquires browsing history information including the address and browsing time of the web page ,
For each of the one or more first users, based on the browsing history information, the web page browsed in a first period that is a period before the generation timing, based on the identified generation timing, A browsing page identifying unit that identifies a web page browsed in the second period, which is a period after the occurrence timing; and
From the plurality of web pages based on changes in the browsing status of the web pages in each of the first period and the second period of each of the one or more first users identified by the browsing page identification unit, A related page specifying unit for specifying an event related page as a web page related to the life event,
Program to function as.

Images