JP2012208904A - Information processing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To estimate interests of a user with high accuracy while protecting privacy of the user of an information processing terminal connected to a network.SOLUTION: This invention assumes a community represented by a social network or a service with a registration system etc. and supposes that services are available. Action histories performed in a terminal of a user are transmitted to a server, and the server grasps the action histories of the user and returns ads which fit the action histories to the terminal of the user.

Description

  The present invention relates to an information processing system.

  Conventionally, until the Internet has become popular, advertisements for products and services provided to consumers by companies and advertisements using mass media such as posters, pamphlets, and various televisions, newspapers, magazines, etc., have been conducted. (Mass marketing). In such mass marketing, it is common for advertisements to be placed non-selectively even for consumers who do not actually have any needs or interest in the product or the like and consequently do not take any purchasing action. there were. For this reason, the cost of these advertising advertisements is enormous, and the advertising advertisements that have been spent on enormous costs are inefficient because it is difficult to lead to sales.

  However, nowadays, due to the spread of the Internet and the advancement of advanced information technology, it is possible to collect and analyze consumer preferences, interests, purchase histories, etc., which were practically impossible until now, through computers and networks. It has become possible. As a result, it is possible to narrow down the target of promotional advertisements through consumer stratification, consumer targeting, etc., and to effectively carry out promotional advertisements for products that match preferences and interests to specific consumers It became possible (direct marketing). Such direct marketing offers great benefits for consumers because it provides products that consumers need, are interested in, match consumer preferences, and are willing to buy. I can say that.

  In direct marketing using the Internet, the web page and content category viewed by the consumer are stored, the consumer's profile (preference / hobby) is estimated, and the consumer's current web page is viewed by the consumer. An example is a so-called context advertisement that displays advertisements that match hobbies and preferences. In direct marketing using the Internet, in order to provide products that meet the more detailed needs of consumers, for example, consumer behavior information on the Internet using personal information collection technology such as cookies and adware. (Web browsing history, e-mail transmission / reception history, SNS (Social Networking Site) transmission browsing history, etc.) are collected and stored, and consumers are interested in their products by registering their personal information through their websites. In many cases, advertisements such as products are performed using personal information.

  By the way, for example, Patent Document 1 discloses a technology that enables distribution of appropriate advertisements corresponding to a user's life history while protecting privacy. In this technology, the user terminal separates the user's life history into public life history information and private life history information, and provides only the public life history information to the advertisement distribution server. Then, the advertisement distribution server sorts the advertisement data including the public life history information in the advertisement distribution condition. In addition, the advertisement distribution server estimates public life history information strongly associated with a predetermined advertisement distribution condition corresponding to the non-public life history information, and preferentially uses advertisement data including the estimated public life history information in the advertisement distribution condition. Deliver to. When the user terminal downloads the advertisement data, the user terminal collates the private life history information with the advertisement data, and displays the advertisement distribution condition that matches the private life history information.

JP 2010-123011 A

  Certainly, in order to effectively provide consumers with products that meet more detailed needs, a lot of detailed information about consumers such as personal information of consumers and behavior history information on the Internet is required. . On the other hand, considering the recent increase in privacy protection and the leakage of personal information, which has become a social issue, detailed information about these consumers must be handled with caution. For example, personal information collected by personal information collection technology such as cookies and adware is not collected with the consent of the individual in most cases, and may become a problem from the viewpoint of privacy protection in the future. That is, when the advertisement distribution side acquires personal information, it must strictly manage the information, which requires a large management cost.

There are several privacy protection methods, but in the invention disclosed in Patent Document 1, only personal information permitted by the user is disclosed. That is, the user's personal information (life history information) is separated into public information (information that may be disclosed) and non-public information (information that is not disclosed), and only the public information is disclosed to the advertisement distribution side. On the server side, an advertisement to be distributed is determined based on the public information. However, as for the action history information on the Internet as personal information of the user, what kind of website was browsed (what keywords are included in the website), what kind of contents were used for sending and receiving emails (What keywords are included in the e-mail), and it is very difficult and unrealistic to separate all of this information into public information and non-public information on the user side. In that case, it is natural that only the information that the user can clearly disclose is disclosed, and if it is originally disclosed, the appropriate information is not disclosed.
Here, it is assumed that only information that each user can clearly disclose can be disclosed. Then, it is convenient if an advertisement to be distributed to a certain user can be estimated based on the public information of other items of other users whose public information is similar. However, in that case, it becomes a problem how to calculate the similarity of public information and how to estimate.

  On the other hand, as another method of protecting the privacy, it is conceivable to limit the personal information to be disclosed to information that cannot pinpoint the user. In other words, it is an idea (called anonymization) that it is not necessary to distinguish from many other users. The degree of privacy protection (called anonymity) becomes so strong that it cannot be distinguished from more users. For example, when a fan of a sports player who has caused a scandal searches for information on that player on the Internet, the number of users to be searched is completely different before and after the scandal is discovered. When the same keyword (for example, the name of the player) is used, the fan can be identified more when the search is made before the scandal is detected, and the tendency is more remarkable as the player is not famous. On the other hand, if the keyword is the name of a sport played by the player (baseball, soccer, etc.) or the team name of the player, it is more difficult to specify the user than searching using the player's name. (There are more users searching by sport name or team name). In this way, it is effective as a privacy protection method that the action history information on the Internet possessed by the user is information that cannot be pinpointed by the user, and this processing is further performed on the user side. It is preferable from the viewpoint of usability to be performed on the device or system side.

It can be determined whether or not it is difficult for the action history information of a certain user to identify the user by comparing with the action history transmitted by another user to the server.
However, as the community grows, the amount of calculation becomes too large in comparison with all users, so it is natural to divide into groups of an appropriate size. However, in this case, depending on the group to which the user belongs, there is a problem that becomes an inefficient state in which it is difficult to specify the user unnecessarily, and the targeting cannot be effectively performed.

  Therefore, an object of the present invention is to enable the user's interest to be accurately estimated while protecting the privacy of the user of the information processing terminal connected to the network.

  In order to solve the above problem, the invention according to claim 1 is an information processing system including a plurality of information processing terminals grouped and a server device that transmits distribution data to the information processing terminals, The server device includes: a server-side receiving unit that receives the action history transmitted by the information processing device; a group ID acquisition unit that acquires a group ID of the group including the information processing device that has transmitted the behavior history; Storage means for storing distribution data associated with a plurality of behavior histories conceptualized in the above, distribution data associated with the behavior histories in a group including the information processing terminal that has transmitted the behavior histories, Group I of a group including a plurality of action histories associated with distribution data and an information processing apparatus that transmits the distribution data for each of the plurality of action histories A server-side transmission means for transmitting D, and the information processing apparatus includes: a terminal-side transmission means for transmitting a conceptualized action history at an arbitrary stage among user action histories; and the server apparatus It comprises receiving means on the terminal side for receiving the transmitted distribution data, and group selection means for selecting a new group from a plurality of behavior histories and the group ID included in the distribution data.

  According to a second aspect of the present invention, in the first aspect of the invention, the information processing terminal further includes a grading means for stepwisely classifying the user's behavior history, and the storage means of the server device , Storing the distribution data and a plurality of concepts layered in stages, and the transmission means on the server side associates the concept with a group including the information processing terminal that transmitted the concept Transmission data, a plurality of concepts associated with the distribution data, and a group ID and the number of information processing terminals of the group including the information processing apparatus that has transmitted the distribution data for each of the plurality of concepts. The group selecting means selects a new group from a plurality of concepts associated with the distribution data, the group ID, and the number of information processing terminals.

  The invention according to claim 3 is the invention according to claim 2, wherein the action history before the conceptualization is compared with the hierarchical action history received from the server, and is more consistent than the group currently belonging It is characterized in that a strong group is a candidate for a new group.

  According to a fourth aspect of the invention, there is provided the storage device according to the third aspect of the invention, comprising storage means for associating and storing the group ID of the candidate of the newly belonging group and the number of times of candidate determination, and a predetermined threshold value It belongs to the group determined as a candidate as described above.

  The invention according to claim 5 is the invention according to claim 4, wherein the storage contents of the storage means are cleared at a timing belonging to a new group.

  The invention according to claim 6 is the invention according to claim 5, wherein the degree of coincidence obtained by comparing the action history before the high-level conception with the hierarchical action history received from the server for the newly belonging group candidate is calculated. The calculation is performed by the number of matching hierarchies.

  The invention according to claim 7 is the invention according to claim 6, further comprising storage means for storing the group ID of the candidate of the newly belonging group in association with the cumulative value of the matching degree, Is newly belonging to a group having a predetermined threshold value or more.

  According to the present invention, it is possible to accurately estimate the user's interest while protecting the privacy of the user of the information processing apparatus connected to the network.

It is a conceptual diagram which shows one Embodiment of the information processing system which concerns on this invention. FIG. 2 is an example of a block diagram illustrating an example of terminal software used in the information processing system illustrated in FIG. 1. It is an example of the block diagram of the software of a server and other terminals shown in FIG. In the information processing system shown in FIG. 1, it is an example of stepwise implementation of privacy information sent from a terminal to a server. In the information processing system shown in FIG. 1, it is an example of the detail of the stage 3 of the privacy information sent from a terminal to a server.

  The present invention assumes a community represented by a social network, a registration service, and the like, and assumes that the service can be received, and sends a history of actions performed on the user's terminal to the server. This includes grasping the history and returning an advertisement suitable for the action history to the user's terminal.

[Embodiment 1]
<Configuration>
FIG. 1 is a conceptual diagram showing an embodiment of an information processing system according to the present invention.
The system shown in FIG. 1 includes a network 5, client devices (hereinafter referred to as terminals) 1a to 1n as a plurality of information processing devices connected to the network 5, a server device 2 connected to the network 5, and a server. A DB (Data Base) 3 connected to the device 2 is provided.
Each of the terminals 1a to 1n has a function of receiving information from the server based on the action history that has been conceptualized, and determining whether to display the received information based on the action history before being conceptualized. PC.
The DB 3 is means for storing information (for example, advertisement information) to be sent to each of the terminals 1a to 1n, and includes, for example, an HDD (Hard Disc Drive).

Comparing the behavior history of many users of other terminals 18 with the behavior history of the user who is the user of terminal 1, the network load and the CPU (Central Processing Unit) load on the terminal 1 side are large and often not realistic. . For this reason, it is natural to perform grouping of an appropriate size.
Distribution data such as advertisements to be distributed to a certain user when estimating public information based on public information of other items of other users in the group, as well as behavior history that should be disclosed in consideration of anonymity, When controlling based on the action history published by other users in the group, it is important to group so that users whose action history is similar to each other belong to the same group It is.
In the initial stage, users of each terminal can only belong to a random group, but in that case, there are few public behavior histories similar to users of other terminals.
In this case, since the estimation based on the public information has low similarity that is the basis of the estimation, the estimation accuracy is deteriorated.
In addition, the behavior history control transmitted according to anonymity is in an excessively high level of abstraction, and an inefficient situation continues. This is because in a stepwise action history transmission with random grouping, even if a certain number of users are performing actions, rare actions below a certain percentage are only counted for a certain user in the random group count. This is because there is a high possibility of action and excessive privacy protection.

  Therefore, if you want to get more accurate advertisements by moving to another group with a behavioral history similar to your own behavior, your behavior as a user of your device is largely biased among the groups to which you belong. Propose a method to determine whether to move to a group.

In the present embodiment, the group to which the terminal 1 of the user belongs is replaced based on the personal information sent by each user to the server 2.
At first, it does not have to belong to a group, all of them are in a side-by-side state, and finally may be divided into a plurality of groups.
The case where it is divided into a plurality of groups will be described below.

The action history having a large bias within the group as described above is obtained by notifying the server 2 side (service side) of the action history in which information embodied from the action history tree is notified to the server 2 side (service side) and the group ID tree information (service side). ) To match the user's own action history, and an appropriate group can be found according to the confidential level of the user of the terminal 1 (how many users he / she wants to be confused with).
This tree information may be in a form in which the information received from the server includes a group ID of a group that has transmitted a lower conceptualized action history and a tree structure of a correspondence relationship between hierarchical action histories.

[Constitution]
FIG. 3 is an example of a block diagram of software of the server and other terminals shown in FIG.
The server 2 shown in FIG. 3 includes behavior information receiving means 21, behavior information storage means 22 for each group, advertisement information generating means 23, and advertisement information transmitting means 24.

  The behavior information receiving means 21 is a means for receiving the signal “rice stamp *” from the terminals 1a to 1n shown in FIG.

  The behavior information storage means 22 for each group is means for storing behavior information of a plurality of groups.

  The advertisement information generation means 23 is means for generating advertisement information for each group (such as a group to which the terminal 1 belongs) from the server 2 side. This advertisement information may include information on a tree structure of a correspondence relationship between a group ID of a group that has transmitted a lower conceptualized action history and a hierarchical action history. Separately, similar information may be obtained from the terminal 1 by inquiring the server 2.

  The advertisement information transmitting means 24 is means for transmitting appropriate advertisement information to each group (such as a group to which the terminal 1 belongs) from the server 2 side.

  The advertisement information is transmitted from the advertisement information transmitting means 24 of the server 2 to the terminal group including the terminal 1.

  The terminal group including the terminal 1 includes an advertising means receiving means 25, a self group evaluating means 26, and a new group determining means 27.

  The advertisement means receiving means 25 is means for receiving advertisement information from the server 2 side.

  The own group evaluation means 26 is a means for evaluating the group to which the user terminal 1 belongs. “Evaluation” means that the group to which the terminal 1 of the user belongs determines the significance and value for the user of the terminal 1.

  The new group determination means 7 is a means for determining whether or not a new group candidate transitioning from the group to which the user terminal 1 belongs has significance and value for the user of the terminal 1.

<Operation>
First, the user's terminal 1 first notifies the server 2 side (service side) of the abstracted action history (this action history notification is also disclosed to other users' terminals in the group to which the terminal 1 belongs. .)
Next, when a certain number of other users in the group to which the terminal 1 belongs transmits the same behavior history as the abstracted behavior history transmitted from the user terminal 1 to the server 2 side, the behavior previously transmitted An action history embodied in stages is transmitted in a form related to the history information.
By embodying in stages as described above, the action history can be notified to the server 2 side (service side) in an appropriately abstracted state.
When the advertisement information is distributed from the server 2 to the group including the terminal 1, data of a tree structure of a group ID corresponding to an action history embodied from the corresponding action history and an action is given.

Hereinafter, a description will be given with reference to FIGS.
FIG. 5 is an example of detailing of stage 3 of the privacy information sent from the terminal to the server in the information processing system shown in FIG.

In FIG. 5, it is assumed that the action history transmitted from the user terminal 1 to the server 2 side is stage 3, that is, the stage of “watching TV now” from the terminal 1. The group to which the terminal 1 belongs is G19. Behavior information “watching TV now” is distributed to the other terminals 18 and the server 2 in G19.
In this stage, it is assumed that the advertisement related to stage 3 distributed from the server 2 to G19 including the terminal 1 includes tree structure information as shown in FIG. Referring to this information, it can be seen that the total number of groups is 19, including G19 to which the user terminal 1 belongs. Further, the number of users included in these 19 groups may be displayed on the display unit of the terminal 1. This display indicates to the user that it is not specified.

If the action history of the user of the terminal 1 is “I am watching Fujiyama TV now”, G1, 2, 3, G4, 5, 6, 7, G8, 9, 10, G11 Be a candidate. On the other hand, G12, 13, 14, 15, 16, 17, and 18 are not candidates for moving to new ones because they become inconsistent in one-step refinement. And the user of the terminal 1 can select the group which should be changed by comparing the number of members of each group with the user's desire for anonymity (how many groups would he like).
G1, 2 and 3 have the same action information at the lowest level of “I am watching Fujiyama TV now”. G, 1, 2, 3, 4, 5, 6, and 7 have the same behavioral information at the level of “I am watching Fuji TV, Hinode TV, or Nippon TV today”. G1,2,3,4,5,6,7,8,9,10,11 coincides with the level of "I am watching live broadcasting now". However, G4, 5 and 6 do not have the same behavior information at the level of “I am watching TV at sunrise now”, and G9 and 10 have the same behavior information at the level of “I am watching terrestrial broadcasting”. Not doing so is not a good candidate compared to G1,2,3. Since G7 does not reveal the lower-level action information, G8 does not reveal the lower two, and G11 does not reveal the lower three, these groups are good candidates if privacy is a priority.

  In the present embodiment, when the advertisement information is generated, the server side also has tree structure information and can notify that there is a better group for the terminal 1.

  The number of cases where the behavior information of other users is the same may be counted, or may be notified from the server 2 or the other terminal 18.

  The above-described embodiment shows an example of a preferred embodiment of the present invention, and the present invention is not limited thereto, and various modifications can be made without departing from the scope of the invention. is there.

1, 1a to 1n terminal (client)
2 Server (site)
3 DB
DESCRIPTION OF SYMBOLS 5 Network 11 Behavior information collection means 12 Behavior information staging means 13 Behavior information transmission means 14 Other person's behavior information reception means 15 Self-rareness unusualness judgment means 16 Detailed staging instruction means 18 Other terminal 21 Behavior information reception means 22 Action information storage means for each group 23 Advertisement information generation means 24 Advertisement information transmission means 25 Advertisement information reception means 26 Own group evaluation means 27 New group determination means

Claims (7)

An information processing system including a plurality of information processing terminals grouped and a server device that transmits distribution data to the information processing terminals,
The server device includes a server-side receiving unit that receives an action history transmitted by the information processing device, a group ID acquisition unit that acquires a group ID of a group including the information processing device that has transmitted the behavior history,
Storage means for storing distribution data in which a plurality of hierarchically conceptualized action histories are associated;
Distribution data associated with the action history, a plurality of action histories associated with the distribution data, and a distribution for each of the plurality of action histories to a group including the information processing terminal that has transmitted the action history Server-side transmission means for transmitting the group ID of the group including the information processing apparatus that transmitted the data,
The information processing apparatus includes a terminal-side transmission unit that transmits a conceptualized behavior history at an arbitrary stage of a user's behavior history;
Receiving means on the terminal side for receiving distribution data transmitted from the server device;
An information processing system comprising: group selection means for selecting a new group from a plurality of action histories included in the distribution data and the group ID. The information processing terminal further includes staging means for conceptualizing a user's action history step by step,
The storage means of the server device stores the distribution data in association with a plurality of concepts layered in stages,
The server-side transmission means includes a group including the information processing terminal that transmitted the concept, distribution data associated with the concept, a plurality of concepts associated with the distribution data, and the plurality of concepts Sending the group ID and the number of information processing terminals of the group including the information processing device that transmitted the distribution data for each,
The information according to claim 1, wherein the group selection unit selects a new group from a plurality of concepts associated with the distribution data, the group ID, and the number of the information processing terminals. Processing system.   The behavior history before the superordinate conception is compared with the hierarchical behavior history received from the server, and a group having a higher degree of matching than the currently belonging group is set as a new group candidate. Item 3. The information processing system according to Item 2.   4. A storage means for storing the group ID of the candidate of the newly belonging group in association with the number of times the candidate has been determined, and newly belonging to a group determined to be more than a predetermined threshold. Information processing system described in 1.   5. The information processing system according to claim 4, wherein the storage content of the storage means is cleared at a timing belonging to a new group.   6. The degree of coincidence obtained by comparing a behavior history before conceptualization with a group candidate newly belonging to a hierarchical behavior history received from the server is calculated as the number of matching layers. Information processing system.   Storage means for storing the group ID of the candidate of the newly belonging group in association with the cumulative value of the matching degree, and newly belonging to a group having a matching degree cumulative value equal to or greater than a predetermined threshold value The information processing system according to claim 6.

Images