JP2002041408A - Direct advertising email system - Google Patents

Abstract

(57) [Summary] An object of the present invention is to provide an advertisement electronic mail system using a data mining process with a good matching ratio. An advertising e-mail system according to the present invention includes:
In a membership-based information database that can be searched for from a keyword, an information advertisement distribution destination corresponding to the advertisement need is obtained from the member's reading history of the information of the information database corresponding to the attribute of the member and specific advertisement needs. Is determined, the advertisement electronic mail in the field of interest of the member is distributed to the member of the determined distribution destination. The information database is a specialized article information database relating to a technology in a specific field.
Use for about a month.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for advertising on the Internet.
The present invention relates to a direct advertisement e-mail system for narrowing down advertisement e-mail distribution destinations from a base.

[0002]

2. Description of the Related Art One of the leading marketing methods using the Internet is "direct advertisement electronic mail". This is a method of delivering advertisement messages by e-mail. E-mail has become more frequently used recently because it requires less printing and mailing costs than direct mail, and can be made cheaper with only distribution costs.
However, as a result, a large number of advertisement mails that are not of interest to them are distributed, and there is an adverse effect that the mails are annoying to users (spam mails).

[0003] In a portal site having a large number of users, in order to maintain the trust of users, it is necessary to avoid the above-mentioned annoying advertisement mail. Therefore, how to distribute advertisement information that suits the user's interest has become an important issue.

Heretofore, there have been proposed several schemes for advertising distribution that match the interests of users as described below. a. Opt-in e-mail Opt-in e-mail is used to deliver advertising e-mails to registered users. The user selects a category (field or theme) of his / her interest on the Web in advance, Is for distributing product information belonging to the category as an advertisement e-mail.

[0005] The opt-in mail is characterized in that it is a kind of "permission marketing" for marketing with the permission of the consumer. However, it is difficult to update the opt-in mail so that the preference information of the user registered in advance is not obsolescent.

For example, assuming business use, a user usually changes sections about once every two years.
After the transfer, I want information in my new section, but no longer need information from the old section. However, it is troublesome to re-add a field of interest to the site every time a user moves a section, so it is not usually performed. Therefore, the advertising mail sent thereafter becomes obsolete.

There is also a disadvantage that it is difficult to deal with the obsolescence of keywords used for categorization. In particular, in the field of computers and the Internet, where new devices and software are evolving on a daily basis, updating this keyword is troublesome and time-consuming.

B. Recommendation Engine In order to increase the effectiveness and efficiency of advertising information dynamically presented to the user, it is an information provision method based on preset rules and user's browsing behavior, using cookies etc., for example, It is intended to present related information (article B, advertisement C) to the user who browsed article A.

In practice, a user who has purchased a product A (for example, a printer) through electronic commerce or the like is offered a product B (for example, a toner cartridge). When the number of products to be handled is several hundred thousand to several million, it takes a lot of trouble to set rules, and there is a disadvantage that targeting cannot be performed if prediction cannot be made in advance.

[0010] In addition, this method often uses a cookie, and in this case, it is assumed that the user releases the cookie. c. Automatic analysis-type recommendation engine Recently, as an improvement of the above-mentioned b, the access history of the user is analyzed to infer the preference of each user, and the optimum product or advertisement is automatically proposed (recommended). Software is being developed. Specifically, the tendency of access, such as the commodities that the user is accessing with difficulty and what kind of content has been accessed recently, is statistically analyzed based on the access history. Next, the user's access tendency is compared with that of another user, and a user having a similar access tendency is extracted. Then, it is a mechanism for sequentially proposing products and contents that are accessed by the extracted user but not yet accessed by the user to be analyzed. The automatic analysis type recommendation engine comprises: a. Eliminates the need for the user of the opt-in mail to input a keyword, and b. The recommendation engine is superior to both in that it eliminates the hassle of entering rules on the site side of the recommendation engine.

However, also in this case, it is complicated to specify parameters for analyzing the access tendency and to perform statistical processing using the parameters, and it takes time to extract users having similar access tendencies. The analysis is difficult when there are one million users. Also, in this case, since the access history of the user is obtained by using a cookie, there is a drawback that the function does not function unless the cookie of the user is released.

[0012]

SUMMARY OF THE INVENTION The present invention relates to an improved version of the above automatic analysis type recommendation engine.
It is another object of the present invention to provide an advertising mail system using a data mining process with a good matching ratio.

[0013]

SUMMARY OF THE INVENTION The present invention is based on the premise that a membership-based search information database exists, and meets the special needs of advertisers based on the reading history of the user in the information database. It is possible to present an advertisement e-mail that is fine-grained and has a high matching ratio.

That is, according to the present invention, in the membership-based information database searchable from a keyword, the information data / data corresponding to the attribute of the member and specific advertising needs is provided.
By selecting the distribution destination of the advertising e-mail corresponding to the advertising needs from the reading history of the member to the base information, the advertising e-mail of the field of interest of the member to the member of the selected distribution destination is selected. Is a direct advertisement e-mail system.

Further, in the system, a keyword in the information database corresponding to the specific advertising need is selected, the information data including the keyword is extracted, and information corresponding to the advertising need is selected. The selected information is specified as a field of information of interest to the member, and the advertisement electronic mail is delivered to the member who has read the selected information.

In this case, an embodiment in which the information database is an information database relating to a specific specialty is recommended. Further, the reading history is reading history information for a predetermined period within the last one year (for example, about the last one month),
It is preferable that the reading history includes category information to which the keyword of the specific field belongs, and the technical terms are unified as the keyword.

The software used in the direct advertisement e-mail / distribution system or distribution method of the present invention is a member-based information database which can be searched from a keyword, and a reading history data base for the member's information database. A process for selecting a category and a keyword for narrowing down an advertisement distribution destination corresponding to a specific advertising need, and a process for selecting information corresponding to the specific advertising need from the category and the keyword for narrowing down. Reading the reading history of the member from the reading history database to the information in the information database corresponding to the specific advertising needs, and determining an advertisement distribution destination corresponding to the advertising needs; Direct, consisting of first delivering advertising emails that address specific advertising needs Characterized in that it comprises a tell distribution electronic program for e-mail, it is stored in the electronic storage medium.

Still further, the software has an attribute data base of the member, the number of times the member has recently accessed the information database is a predetermined number or more, and the advertisement distribution destination is narrowed down. It is preferable to include a step of identifying whether or not the information for reading has been read a predetermined number of times, and a step of checking whether the member attribute matches the member attribute for narrowing down the distribution destination.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 shows an overall view of a network used in the present invention. In FIG. 1, reference numeral 1 denotes the Internet, and information devices such as a general user's personal computer 2 and a mobile phone 3 are connected to a WWW server 5. The server side is W
In addition to the WW server 5, an information processing server 4, a member DB 6, an access history DB 7, a specialized article information DB 8, a company (advertiser) DB 9, and the like are connected inside the firewall 10.

FIG. 2 is a flowchart showing a series of flows of the direct advertisement electronic mail system of the present invention. The specialized article information database 8 in FIG. 1 is a member-based database (hereinafter, the database is referred to as a DB), and a user first performs a member registration (step 2).
1). Member registration includes your company, department, date of birth,
Basic attributes such as posts are input. These are stored in the member DB of FIG. FIG. 5 shows an example of the member DB. 5 is raw data according to the time series at the time of registration, but in the case of reporting to an advertiser or the like, the data is output in a simplified form of FIG. .

At the time of the above registration, the following procedure is prepared in order to prevent spoofing registration (step 27). Fig. 8-
Description will be made with reference to the screen example of FIG. 10. 1) The user first performs temporary registration from the form on the WEB site. (FIG. 8) 2) A dedicated URL for the official registration is sent by e-mail to the above temporarily registered address. (Fig. 9) 3) Complete the official registration from the official registration form. (FIG. 10)
(SSL compatible. However, SSL; SecuritySockets Laye
r) 4) A registration confirmation email is sent to the user.

Although the E-mail address is used for the user ID, in the above 2) and 4), the notification of the temporary registration or the formal registration is sent to the E-mail address of the user, so the E-mail address of another person is used.
Spoofing using the il address is prevented.

Next, the reading action when the user accesses the specialized DB is captured (step 22). The details of the specialized DB and the method of capturing the above reading behavior will be described later. Although a cookie is used to save the trouble of logging in the user, the reading behavior can be grasped by the login data without opening the cookie because of the access-controlled member DB.

Next, a user who is likely to be interested in the contents of the advertisement mail is extracted (step 23), the contents of distribution are optimized (step 24), and the direct advertisement electronic mail is delivered (step 25). Advertisement e-mail distribution allows a delay from the occurrence of an advertising need to the transmission of information (advertisement e-mail distribution). This is a point different from a dynamic service such as a conventional recommendation engine. The present invention thereby employs the above procedure and solves the conventional disadvantages.

This distribution is performed in the following procedure. 1) The user's e-mail address (C
SV format. However, CSV: Comma Seperated Value)
Is registered to the mail delivery server by ftp (file transfer).

2) At the same time, the body of the advertisement mail to be distributed is registered in the server. 3) The advertisement mail is delivered at an arbitrary reserved time.
After the delivery of the advertisement mail, a report of the clicker is created as necessary (step 26).

4) In the mail delivered in step 25, other WE
When a link (clickable URL) to the site B is described, it is possible to grasp not only the number of clicks by the user but also which user clicked through the CGI (32) in FIG. . However, reporting will be performed to the extent that individuals cannot be identified. CG
I is an abbreviation for Common Gateway Interface.

Hereinafter, the specialized article information DB will be described in detail. The information on specialized articles is, for example, information created by the same applicant, extracted from various specialized journals, newspaper articles, and the like, and stored as an information DB. The contents are published on the WEB. I have. The specialized article information DB has the following features.

A. Highly specialized articles that are updated daily are accumulated, and scripts and terms are unified. b. As shown in FIG. 12, highly versatile themes are categorized and provided as theme-specific sites. Of the tens of thousands of articles, those that match the above categories (themes) are located under the directory of each site.

C. Therefore, it is possible to set an extraction condition with less ambiguity from a technical article and technical terms in the article. For example, NEOnline and a specialized article information DB provided by the same applicant are described in detail below.

NEOnline is a comprehensive information site for electronics, and the main target audience is R & D
He is an electronic engineer in the department. There are multiple specialized article information DBs used for NEOnline (sites by theme)
Consists of A plurality of sites (thematic sites) are registered, but if you register for one site, you can use other sites.

FIG. 11 shows a screen example of NEOnline. There are 5.8 million monthly page views, 20 satellite sites, and about 220,000 registered users. FIG. 3 shows the overall system and the data flow associated therewith. The upper part of the figure shows the flow of access of the user to the specialized article information DB. When the user 31 accesses the specialized article information DB 36 from the website 32, the user 31 is first authenticated.
When the authentication is NG, the user is registered from the user registration page 33 and registered in the user information DB 34 and the authentication DB 35. The specialized article information DB 36 is divided into 20 themes and other satellite sites (theme-specific sites), and the user searches and reads the specialized article information DB as needed. Since the access log of the reading is stored in the access history DB as shown in FIG. 7, when the mining rule of the distribution of the advertisement electronic mail (hereinafter, referred to as special mail) is set (38), the access log is read. The log 37 is screened and the basic attribute information of the user is taken in (15), data mining is performed, and an E-mail address list 39 of the user that matches the mining rule is created. Hereinafter, an E-mail (advertising electronic mail) 14 is distributed to each user using the mail distribution system 13.

Next, the clicker counter C in FIG.
The GI 32 will be described. FIG. 13 shows a display example of a direct advertisement electronic mail (special mail) of the present invention. The URL (CG
When the user jumps by kicking the URL (I URL and parameter) 51, the user's E-mail address is picked up from the user's cookie. That is, as shown in the spoofing prevention process described above, in the system of the present invention, the user ID is
Since the e-mail address is used and the ID written in the cookie is the e-mail address itself,
Write it as g. In the access history DB,
1. Date and time; User name obtained from user cookie,
3. 3. Password obtained from user cookie; 4. Satellite site name; 5. Advertising name obtained from datatxt, 6. Browser type, For example, an IP address. FIG. 14 shows an example of the operation.
Shown in

An embodiment of capturing the reading action when the user accesses the specialized article information DB and distributing the direct advertisement electronic mail (special mail) of the present invention will be described below with reference to FIG.

For example, an advertiser may say “Young (40 years old or younger)
V to be implemented in Osaka
I would like to provide seminar information on LIW (VeryLong Instruction Word) processor by e-mail. ", Select" Embedded Systems "as the thematic site category and enter the most recent (eg, last month) articles (α, β,
γ) (step 44) and recently (eg, in the last month) accessed “embedded systems” as the thematic site category above, eg, more than 10 times, and received recent articles (α, β, γ). ) Extract the user who read both of them (step 45).

Next, a user who is under 40 years old and lives in Osaka is extracted from the basic user attributes (step 46), and the special mail is delivered. Examples of the keywords to be picked include the following. a. Elemental technology terms: Bluetooth, Verilog-HDL, ADSL b. Company name: Sony, Fujitsu, NEC c. Conference / Exhibition: ISSCC, COMDEX, EDA
Technofair d. Product name (set): PlayStation2, Tau, DVD recorder e. Product name (device): Crusoe, LED drive IC, A
MDK6-2 + In addition, the name of the development method, equipment (investment), personnel affairs (personal name), etc. are also listed.

A. When the user of the opt-in mail inputs a keyword, or b. Recommendations
In any case of inputting rules on the engine site side, inputting a keyword of a considerably lower concept such as the above-mentioned "VLIW processor" in advance requires a number of thousands to tens of thousands of units. Impossible.

The above c. In the case of the automatic analysis type recommendation engine, access trends such as the products that the user is accessing with difficulty and what kind of content they have recently accessed are statistically analyzed based on the access history, and then Compare the access tendency of the user with the access tendency of other users, extract users with similar access tendency, products and contents that the extracted user has accessed but the user to be analyzed has not accessed yet The purpose is different from that of narrowing down the keyword by a considerably lower concept such as "VLIW processor" as in this request. The automatic analysis type recommendation engine cannot meet the above demand.

In other words, the feature of the present invention lies in that, based on the specialized article information DB, effective distribution destinations are narrowed down according to the request of the advertiser for users who have accessed the specialized article information DB. is there. In addition, it is preferable that the technical article information DB is used for unifying characters and terms. This is because the unification of scripts and terms can eliminate omission of search at the time of narrowing down. However, scripts and terms have not been unified,
If some omission of retrieval is allowed in the information database, the special mail distribution system of the present invention can be applied. Similarly, it is preferable that the specialized article DB is categorized and provided (as a theme-specific site). This is because the categorization improves search efficiency. However, it goes without saying that the special mail distribution system of the present invention can be applied to an information DB that is not categorized.

As described above, by searching the recent user's reading history, it is possible to eliminate the unnecessary delivery of unnecessary advertisement mails after the section is moved. In addition, by limiting the distribution destination according to the number of times of access, it is possible to eliminate users who do not access much, that is, users who are not keen on collecting information. Further, as described above, the basic attributes of the user include the company to which the user belongs, the date of birth, and the like.
Since only a static thing that hardly changes is input, it is limited to those that can be used semi-permanently without updating by the user thereafter. Therefore, it is possible to sufficiently cope with a tendency that the user does not frequently update data after the initial input.

Note that typical examples of such demands include the following simple ones in addition to the above. For example,
"3000 technicians interested in interface related"
Can be much easier than in the above example, because it is only necessary to extract the top 3000 people who have accessed the contents under the "interface" directory under the access control under the highest control. Even if the condition "other than Nikkei BP" is added to this, the affiliated company information may be extracted from the member information DB and matched.

[0043]

According to the present invention, the direct advertisement e-mail
The systems and methods provide significant advantages not possible with conventional advertising emails, such as:

(1) Based on the reading history of the information DB such as the specialized article information DB, the reading history is reversely searched, an article in a field of interest of the user is specified, and the user who has read the article is displayed on the advertisement electronic. By narrowing down the mail delivery destination, it is not necessary to input a term or the like for narrowing down the delivery destination on the user side or the site side. In other words, this is a method for automatically determining a distribution destination that does not require any new input. (Note that, to be precise, the search keyword used when the user read the information database is used as it is. However, actually, the keyword is used for the article search for narrowing down the advertisement distribution destination. (2) The keywords used for narrowing down the distribution destination are:
Since the keyword itself is used for searching the information DB, the number thereof is in the range of tens of thousands to hundreds of thousands, and it is possible to narrow down the target quite finely. In the conventional pre-input type, pre-check input to such a large number of keywords is practically impossible.

(3) As described above, since it is possible to narrow down the target very finely, it is possible to distribute the advertisement mail corresponding to the needs of the advertiser, which cannot be realized by the conventional system, and which is narrowly targeted. .

(4) An access log used for the narrowing down is recent (for example, recent one to three months)
With the use of, it is possible to respond to the latest needs of the user, and it is possible to prevent the advertisement e-mail in a field that is no longer of interest from being distributed as spam e-mail, for example, by moving a section. In the conventional distribution method, it is difficult to input an update corresponding to this need, whereas in the advertisement e-mail of the present invention, the distribution destination is always automatically updated.

(5) It is preferable that the terms of the keywords are unified, since the trouble of inputting is omitted and the omission of narrowing down the target is reduced. (6) In the end, the direct advertisement e-mail system and method of the present invention have a very good matching rate to the user's preference as described above, and therefore have less repulsion from the user than conventional e-mail. In other words, this is an extremely excellent method for distributing an advertisement electronic mail.

[Brief description of the drawings]

FIG. 1 is a diagram showing an overall conceptual diagram of a network system in which the present invention is used.

FIG. 2 is a diagram showing an overall flow of a special mail according to the present invention.

FIG. 3 is a diagram showing a system used for special mail according to the present invention and a data flow in the system.

FIG. 4 is a diagram showing an example of data mining of special mail according to the present invention.

FIG. 5 is a diagram showing the contents of a member DB.

FIG. 6 is a diagram showing a member information output format in the case of reporting.

FIG. 7 is a diagram showing an example of the contents of a user access history DB.

FIG. 8 is a diagram illustrating an example of a display screen (a screen for provisional registration of a user) used in a spoofing registration prevention flow.

FIG. 9 is a diagram illustrating an example of a display screen (temporary registration completion screen) used in the spoofing registration prevention flow.

FIG. 10 is a diagram showing an example of a display screen (final registration screen) used in the spoofing registration prevention flow.

FIG. 11 is a diagram illustrating an example of a display screen of a specialized article information DB.

FIG. 12 is a diagram illustrating the configuration of a specialized article information DB.

FIG. 13 is a diagram showing a display example of a special mail according to the present invention.

FIG. 14 is a diagram illustrating an operation example of a clicker counter CGI.

[Explanation of symbols]

 Reference Signs List 1 network 2 personal computer 4 information processing server 5 WWW server 6 member DB 7 access history DB 8 technical article DB 9 company (advertiser) DB 13 mail distribution system 31 user 32 Web site 36 technical article DB

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] August 28, 2000 (2008.2.
8)

[Procedure amendment 1]

[Document name to be amended] Drawing

[Correction target item name] Fig. 1

[Correction method] Change

[Correction contents]

FIG.

Claims (9)

[Claims] Claims: 1. In a membership information database searchable from a keyword, the advertisement needs are met from the member's reading history to the information of the information database corresponding to the attribute of the member and specific advertising needs. A direct advertisement e-mail system characterized in that by selecting the distribution destination of the advertisement e-mail, the advertisement e-mail in the field of interest of the member is distributed to the member of the selected distribution destination. 2. Selecting a keyword in the information database corresponding to the specific advertising need, extracting the information data including the keyword, selecting information corresponding to the advertising need, and selecting the selected information. 2. The direct advertisement e-mail system according to claim 1, wherein the advertisement e-mail is delivered to the member who has read the selected information by specifying the field of information of interest to the member. 3. The direct advertisement electronic mail according to claim 1, wherein the information database is an information database relating to a specific specialty field.
system. 4. The method according to claim 1, wherein the reading history is recent reading history information.
Direct advertising e-mail system as described in section. 5. The direct advertisement electronic mail system according to claim 1, wherein the reading history includes category information to which a keyword in the specific field belongs. 6. The direct data processing method according to claim 1, wherein each information is described using unified scripts and terms in the information database. Advertising email system. 7. In a membership-based information database that can be searched for from a keyword, the above-mentioned member's attribute and information on the above-mentioned information database corresponding to a specific advertisement need are met from the member's reading history to the above-mentioned advertisement needs. A direct advertisement e-mail distribution method, wherein an advertisement e-mail in a field of interest of the member is distributed to a member of the selected distribution destination by selecting the information advertisement distribution destination. 8. A member-based information database which can be searched from a keyword, and a reading history database for the member's information database, for narrowing down an advertisement distribution destination corresponding to a specific advertising need. A process for selecting a category and a keyword, a process for selecting information corresponding to the specific advertising need from the category and the keyword for narrowing down, and a process for selecting the information corresponding to the specific advertising need from the reading history database. Information data
A direct advertisement electronic system comprising a step of reading the member's reading history into base information, selecting an advertisement distribution destination corresponding to the advertisement needs, and distributing an advertisement e-mail corresponding to the specific advertisement needs to the distribution destination. An electronic recording medium comprising an electronic mail delivery program. 9. Having an attribute database of said members,
A step of identifying whether the number of times the member has recently accessed the information database is a predetermined number of times or more and has read the information for narrowing down the advertisement distribution destination a predetermined number of times, and the member attribute and the distribution 9. The electronic recording medium according to claim 8, further comprising a distribution electronic program for direct advertisement e-mail, comprising a step of checking a match with a member attribute for narrowing down.