JP2003030232A - Method for distributing information in network - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information distributing method in a network by which information can be distributed to an accurate user by analyzing interest of the user. SOLUTION: A server 1 is connected to the Internet 2 and stores information including a photograph of merchandise for sales transmitted from a marker being an information provider and an information parameter representing attributes of the merchandise in an information registration server 11, an evaluation storage server, and a binary data storage server 15. A trader being a user represents a field in which the trader is interested with an interest parameter having the same element as that of the information parameter and stores the interest parameter in a receiver DB 12. The server 1 operates the information parameter and the interest parameter, decides information distribution to the user on the basis of this result and distributes the information from a distribution server 13. The server 1 also feedbacks return mail from the user to the interest parameter.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for distributing information in a network, and more particularly to an opt-in mail that selectively distributes information matched with a user's field of interest in direct mail by electronic mail.

[0002]

2. Description of the Related Art Conventionally, when sending a direct mail by e-mail, a potential user is asked to answer a questionnaire in advance to investigate a field in which the user is interested, and the user is informed based on the result of the investigation. There is known a method called opt-in mail that appropriately delivers direct mail of contents of interest. With this method,
Compared to indiscriminately sending a large number of direct mails, users can receive only direct mails that are limited to their fields of interest, which makes it possible to increase the direct mail browsing rate. There was a merit. On the other hand, a company that sends direct mail is also effective because it does not deliver direct mail to users who are not interested in the information and can certainly deliver direct mail to users who are interested in the information. We were able to deliver efficient direct mail.

[0003]

However, the actual attributes of goods and services include multifaceted elements, and it is difficult to simply classify them based on only the raw materials and the purpose of use. Therefore, there is a problem that even if the category of products / services is forcibly applied to the field in which the user is interested and the direct mail is delivered, it is not always possible to deliver the necessary and sufficient direct mail to the interested person. . It can be said that the user's interest at the time of the survey is reflected in the results of the survey, but the fields in which the user is interested are fluid ones that change momentarily based on the user's environment and the user's own changes. Is. Therefore, if a fixed decision is made based on the result of the survey, the user's interest will gradually deviate.
As a result, there was a problem that we would continue to send direct mails about fields that we were no longer interested in. In addition, if the user is frequently surveyed to respond to changes in the user, not only will the user be troublesome and will not be able to actively cooperate with the user, but also will not be able to view or receive direct mail. There was a problem of fear. Furthermore, if only direct mail replies are analyzed and only those that have replied are the areas of interest, it will be analyzed as if they were just interested in the replies that happened only then, and the target of the user's interest. There was also a problem that was not reflected correctly. On the other hand, if only the result of the questionnaire to the user or the returned direct mail is the field in which the user is interested, the direct mail is not delivered to the field in which the user is newly interested, and the field is changed. There was also a problem that it was fixed and new demand could not be cultivated. The present invention is
In order to solve the above problems, a method of distributing information in a network for analyzing a user's interest in a more realistic form and distributing necessary and sufficient direct mail relating to arbitrary information to an appropriate user. The purpose is to provide.

[0004]

In order to solve the above-mentioned problems, in the information distribution method in the network according to claim 1, the field connected to the network is divided into predetermined elements in the computer connected to the network. Then, the step of storing the interest parameter, which is stored as an interest parameter composed of a combination of the parameters of each element, and the attribute of the information to be distributed are divided into the same elements as the interest parameters, and as an information parameter composed of the combination of the parameters of each element. Distribution is determined by the step of storing the information parameter to be stored, the step of determining whether or not the information can be distributed by calculating based on the interest parameter and the information parameter, and the step of determining the distribution availability. The step of delivering the distributed information over the network. The gist.

In the information distribution method in the network according to this configuration, the interest of the user who receives the information is analyzed as an interest parameter divided into a plurality of elements in a more realistic manner, and the information to be delivered is also the same. By dividing into information parameters divided into elements, the tendency of the user's interest can be grasped more accurately, and the attribute of arbitrary information to be distributed can also be analyzed more accurately, so that it has a tendency closer to the user's interest. Information can be distributed. Therefore, the user can receive only the information that matches his / her interest more, and the person who distributes the information can also receive the information.
There is an effect that information can be more efficiently delivered to necessary and sufficient users without waste.

In the invention described in claim 2, in addition to the configuration of the information distribution method in the network described in claim 1,
In the step of determining whether or not to deliver, the inner product of the parameters of the corresponding elements of the interest parameter and the information parameter is calculated, this value is compared with a predetermined threshold as a reference, and the availability of delivery is determined based on the result. The point is to judge.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to claim 1, the influence of each element of the user's interest and the attribute of information is adjusted in the step of determining whether or not to distribute. Therefore, there is an effect that the degree of coincidence can be accurately determined.

According to the invention described in claim 3, in addition to the configuration of the information distribution method in the network described in claim 2,
The gist of the present invention is to further include the step of setting a distribution strength for uniformly correcting each parameter of the information parameters at the same magnification in the step of determining whether or not to distribute.

In the information distribution method in the network according to this configuration, in addition to the function of the information distribution method in the network according to claim 2, when the distributor wants to distribute to a large number of users, a more appropriate user is selected. Then there is the effect that it can be delivered.

According to the invention described in claim 4, in addition to the configuration of the information distribution method in the network described in any one of claims 1 to 3, each parameter of the interest parameter is uniformly scaled at the request of the user. The gist is that the method further comprises a step of setting the receiving sensitivity to be corrected to.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to any one of claims 1 to 3, the user's interest in the user who wants to receive a large amount of information. There is an effect that it is possible to increase the amount of distribution of information that matches with.

According to the invention described in claim 5, in addition to the configuration of the information distribution method in the network according to any one of claims 1 to 4, from the information determined not to be distributed in the step of determining the distribution permission / inhibition, It is a gist to extract information based on a predetermined extraction method and distribute the information by the distribution step.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to any one of claims 1 to 4, a new range of interest that is not distributed as it is You can distribute information and develop new interests of users,
This has the effect of preventing the contents of distribution from being fixed.

According to the invention described in claim 6, in addition to the configuration of the information distribution method in the network described in claim 5,
The predetermined extraction method is any one of the information parameter, the interest parameter, the distribution intensity, and the reception sensitivity.
The gist is to extract by making a predetermined change to the above numerical values and then comparing with the threshold value.

In the information distribution method in the network according to this configuration, in addition to the function of the information distribution method in the network described in claim 5, the interest of the user and the attribute of the information relatively match among the undistributed information. It has the effect that you can select one and reduce the delivery of completely unrelated mail.

According to the invention described in claim 7, in addition to the configuration of the information distribution method in the network according to claim 5 or 6, the predetermined extraction method includes a step of extracting based on a random number. Use as a summary.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to claim 5 or 6, it is possible to avoid concentration in a specific field when distributing information. There is an effect that can be.

According to the invention described in claim 8, in addition to the configuration of the information distribution method in the network according to any one of claims 1 to 7, in addition to the evaluation shown by the user who has received the information. An information evaluation receiving step of receiving a certain information evaluation from the user; and a first interest parameter modifying step of modifying an interest parameter corresponding to the information parameter based on the information evaluation received in the information evaluation receiving step. The point is to have prepared.

In the information distribution method in the network according to this configuration, in addition to the function of the information distribution method in the network according to any one of claims 1 to 7, the reaction of the user is evaluated by the information evaluation returned by the user as an interest parameter. To be able to respond to changes in the user's interests.

According to the invention described in claim 9, in addition to the configuration of the information distribution method in the network according to any one of claims 1 to 7, the information distributed to the user is transmitted from the user at the time of distribution. The gist of the present invention is to have a second interest parameter correction step of correcting the interest parameter corresponding to the information parameter based on the length of time until the information evaluation is returned.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to any one of claims 1 to 7, the information distributed to the user is received from the user received from the distribution. By modifying the interest parameter according to the length of time until the information evaluation is returned, there is an effect that it is possible to perform distribution that more matches the interest of the user.

According to the tenth aspect of the invention, in addition to the configuration of the information distribution method in the network according to any one of the first to seventh aspects, a part of a plurality of information simultaneously distributed to the user The summary is that a third interest parameter correction step of correcting the interest parameter corresponding to the information parameter of the information for which the information evaluation is not returned is provided when only the information evaluation is returned.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to any one of claims 1 to 7, the information which has returned the information evaluation and the information evaluation are returned. By providing a difference between the missing information and the information parameter, there is an effect that the distribution can be more matched to the interest of the user.

According to the invention of claim 11, in addition to the configuration of the information distribution method in the network according to any one of claims 8 to 10, the correction of the interest parameter includes a reply of information evaluation by the user. The main point is that it is performed sequentially for each time.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to any one of claims 8 to 10,
There is an effect that the information evaluation from the user can be reflected in the interest parameter in real time.

In the invention described in claim 12, in addition to the configuration of the information distribution method in the network according to any one of claims 8 to 10, the modification of the interest parameter is made in response to the information evaluation from the user. The gist of the present invention is to include a reaction pool for temporarily accumulating based data, and including a step of correcting the interest parameter when a predetermined numerical value accumulated in the reaction pool becomes a predetermined threshold value or more.

In the information distribution method in the network according to this configuration, in addition to the function of the information distribution method in the network according to any one of claims 8 to 10,
There is an effect that the interest parameter can be corrected at the stage where the changes are accumulated, without making unnecessary corrections due to a temporary change in the information evaluation of the interest parameter.

According to the invention described in claim 13,
In addition to the configuration of the information distribution method in the network described in (3), the whole or a part of the numerical value accumulated in the reaction pool is gradually reduced or discarded with the passage of time.

In the information distribution method in the network according to this configuration, in addition to the function of the information distribution method in the network according to claim 13, the interest parameter is not unnecessarily modified by a temporary change of the information evaluation,
The effect is that the parameter of interest can be modified by a continuous change.

According to the invention described in claim 14, in addition to the configuration of the information distribution method in the network according to any one of claims 8 to 10, the modification of the interest parameter is made in response to the information evaluation from the user. The gist of the present invention is to provide a reaction pool for temporarily accumulating based data, and to modify the interest parameter based on a predetermined numerical value accumulated in the reaction pool at regular time intervals.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to any one of claims 8 to 10,
There is an effect that it is possible to prevent the information evaluation for a long time from not being fed back to the interest parameter.

According to the invention described in claim 15, in addition to the configuration of the information distribution method in the network according to any one of claims 1 to 14, in addition to the evaluation shown by the user who received the information, Displaying a list on the user's terminal by associating the information evaluation receiving step of receiving a certain information evaluation from the user, the distributed information, and the information evaluation received in the information evaluation receiving step related to the distributed information. And a step of displaying the received information list list.

In the information distribution method in the network according to this configuration, in addition to the operation of the information distribution method in the network according to any one of claims 1 to 14,
The mail received by the user can be associated with the user's own evaluation and displayed in a list on the user's terminal. Therefore, the user can easily organize and utilize the received emails based on the user's own evaluation.

[0034]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of an information distribution method in a network embodying the present invention will be described below with reference to FIGS.
1 will be described.

FIG. 1 is a conceptual diagram showing the entire network to which information is distributed. Here, the network is the Internet 2, and the server 1, which is a server computer connected to each other via the Internet 2,
Maker 3 (3a, 3b) which is a client computer
...), traders 4 (4a, 4b ...), buyers 5 (5
a, 5b ...). Of course, the network is not limited to the Internet, and any type such as an open or closed LAN (Local Area Network) or WAN (Wide Area Network) may be used.

The maker 3 is a manufacturer who mainly produces products. The buyer 5 is a trader who mainly purchases products such as a department store and a specialty retailer.
Is an intermediary company such as an importer / wholesaler that facilitates the distribution of merchandise through the intermediary of the maker 3 and the buyer 5. Maker 3 to Trader 4, Client 4 to Buyer 5
The information is provided via the server 1. Since the method of distributing information from the maker 3 to the trader 4 and the method of distributing information from the trader 4 to the buyer 5 are almost the same, here, the maker 3 is used as the user to distribute the information about the sold product. The case will be mainly described.

FIG. 2 is a conceptual diagram showing the system configuration of the server 1. The server 1 is composed of a plurality of computers forming a well-known server system, and has an information registration server 11, a recipient information DB 12, a distribution server 13, a WEB server 14, a binary data storage device (image) 15, and distribution information for each user. The database 16 and the evaluation storage server 17 are provided.

The information registration server 11 is a storage means for storing the information of the product related to the information to be distributed and the attribute of the product into the same elements as the interest parameters and storing the information parameters as a combination of the parameters of each element. . Here, for example, in the server 1, for any product, it is divided into items such as “wooden products, pottery, interior products, kitchen products, storage products” in three levels of 0, 1, 2 and the degree of association for each item. The information parameter created in accordance with is stored. For example, in the case of "wooden cabinet", "wooden products 2
・ Ceramics 0 ・ Interior supplies 1 ・ Kitchen utensils 0 ・ Storage items 2 ", 2 points for related wooden products and storage items, 1 point for relatively related interior items, 1 point for relatively 0 for thin pottery and kitchen utensils
Create the information parameter by giving the point. With this information parameter, even products that are difficult to classify can be digitized as data that reflects their attributes.

The recipient information DB 12 stores information about the registered trader 4 who is the recipient of the distributed information. Data such as the trader's name, address, electronic mail address, and telephone number is stored in this area.

The distribution server 13 makes a predetermined calculation on the basis of the interest parameter and the information parameter to determine whether or not the information can be distributed, and sends the decided distribution information to the trader 4 by e-mail via the Internet. To deliver.

The WEB server 14 is available at www (world
A wide web server that is a URL specified by the maker 3 or the trader 4 using the http protocol.
Page is displayed in HTML. In addition, CGI (Com
mon gateway interface), for example, when a user registration or information registration procedure is performed, a program of the information registration server 11 or the recipient information DB 12 is activated to perform a predetermined process.

The binary data storage device (image) 15 is
It is a storage means for accumulating image data to be distributed as binary data. It is sent as an attachment to an electronic mail to the trader 4.

The distributed information DB 16 for each user is a storage means for storing the information distributed to the trader 4 for each user. For example, it is used to charge the distribution fee to the maker 3.

The evaluation storage server 17 is a server for storing interest parameters set at the time of registration of the trader 4 and data returned from the trader 4 as information evaluation with respect to the information distributed to the trader 4. The interest parameter is obtained by dividing the field of interest into the same items as the information parameters and expressing the field of interest of the user as a combination of parameters of each item. The interest parameter is set as an initial value by the trader 4 when the trader 4 is registered, transmitted, and stored therein. Further, thereafter, information evaluation data, which is a reaction from the trader 4 with respect to the information transmitted from the server 1, is converted into the reaction pool 17a.
Is temporarily stored, and when a predetermined threshold is exceeded, the interest parameter of the user is fed back and corrected.

The procedure of distributing information from the server 1 thus configured to the trader 4 will be described. Here, FIG. 3 is a flowchart showing an outline of the flow of information distribution in the server 1. First, a user registration process (step 1 (hereinafter abbreviated as S1)) is performed, and then an information registration process is performed (S2). A distribution determination process is performed to determine whether or not distribution is possible for the user who has registered the information for which the information registration process has been performed (S3). When distribution is determined, information distribution processing is performed (S4), and reply reception processing is performed to receive an information evaluation that is a reply to the information (S4).
5). The received contents are analyzed, and the data update process for updating the data related to the user's interest parameters is performed (S
6) If it is finished (S7: YES), the process is finished (END), and if it is not finished (S7: NO), the procedure from S1 is repeated. This user registration process (S1)
Corresponds to the step of storing the interest parameter of the present invention,
The information registration process (S2) corresponds to the step of storing information parameters, the distribution determination process (S3) corresponds to the distribution permission / inhibition step, and the distribution process (S5) corresponds to the distribution step. Further, the data updating process (S6) corresponds to the first interest parameter correction step, the second interest parameter correction step, and the third interest parameter correction step. Further, although each step is arranged in series for the sake of convenience of description, it goes without saying that each processing may be processed in parallel by using the transmission from the user as an event. Hereinafter, each procedure will be described in detail.

First, in the user registration (S1) process, when the trader 4 who is a user requests user registration, the WEB server 14 receives the recipient information DB via the CGI.
The registration program 12 is activated to display the user registration screen P1 on the trader 4. Here, FIG. 4 is an example of the user registration screen P1 displayed on the terminal of the trader 4. On the user registration screen P1, a text input box for inputting the "name" and "mail address" of the trader who is the recipient is displayed, so the name and the email address of the electronic mail are input here.

Further, the interest of the trader is divided into items such as "wooden products, pottery, interior products, kitchen products, storage products", and the items are displayed on the screen in three stages of "0, 1, 2". Drag the slider with the mouse to enter. Here, the interest parameter of this user is generated based on the input result and is stored in the evaluation storage server 17.

Similarly, the receiving sensitivity is set by the slider. Here, the “reception sensitivity” is to increase the amount of information to be received by increasing the sensitivity. In particular,
It has a function of changing the numerical values of all the parameters of interest by multiplying by a constant coefficient to increase or decrease the amount of information to be decided upon distribution decision. The user can obtain a large amount of information by increasing the reception sensitivity, while the user can obtain only a small number of selected information by decreasing the reception sensitivity.

As described above, the user inputs the name, mail address, interest degree, and reception sensitivity, confirms the contents, and if there is a defect, clicks the "Cancel" button and re-inputs.
If there is no problem in the contents, click the “Register” button to transfer the data to the server 1. This is the end of the user registration procedure (FIG. 3: S3).

Next, the procedure of information registration (FIG. 3: S2) will be described. When there is a request from the maker 3 to provide information about the product to be sold, the WEB server 14 sets C
The registration program of the information registration server 12 is activated via the GI to cause the maker 3 to display the information registration screen P2. Here, FIG. 5 is an example of the information registration screen P2 displayed on the terminal of the maker 3. In the information registration screen P2,
"Product name", "Outline", "Specifications", "Price", "Lot"
A text input box for inputting is displayed. Enter the product name, outline, specifications, price and lot from the keyboard. In the "item name", the name of the item to be sold is entered as in "rattan basket (round shape)". In the "Summary", the description of the product is entered as "Handmade varnish finish". "Specifications" include the size of the product,
The material and specifications are 40 cm (diameter) x 60 cm.
(Height) ”is input. In “Price”, the price and the condition are input like “1,500 $ (FOB)”. In "lot", "500" is input as the quantity of the minimum transaction unit of the product.

Further, the attributes of the products are divided into items such as "wooden products, pottery, interior products, kitchen products, storage products", and displayed on the screen for each item in three levels of "0, 1, 2". The slider is dragged with the mouse to input. Of course, the stages may be subdivided. The information parameter of this product is generated by the result input here and is stored in the information registration server 11.

Similarly, the "delivery strength" is set by dragging the slider displayed on the screen. Here, the “delivery strength” is to increase the number of mails to be delivered by increasing the transmission strength. Specifically, the judgment value is changed by increasing or decreasing the numerical value of the entire information parameter by multiplying it by a constant coefficient, and the number of emails that are compared with the threshold when determining delivery and determined to be delivered is reduced or increased. Have the function to The manufacturer 3 can transmit a large amount of information by increasing the distribution strength, while the manufacturer 3 can provide information to only a selected small number by decreasing the distribution strength. For example, there is a merit in the price of the product,
In addition, when there is a need for a wide range of users and a large amount of sales can be secured, it is conceivable to increase the distribution strength, and conversely to weaken the distribution strength for a small number of limited production sales. The information strength distribution is a relative value within the same maker for controlling how strongly the product as a maker wants to be notified compared to other products. Even if the distribution is the strongest, not all parameters that are absolute values will be maximized.

If there is a photograph of the product, the file name "***. Jpeg" of the photograph of the JPEG format recorded in the terminal of the maker 3 is designated,
Click the "Attach" button. Then, it is stored in the binary data storage device (image) 15 for the server 1 and attached to the transmission file at the time of distribution.

As described above, the product name, outline, specifications, price,
Enter the lot, information parameter, distribution intensity, and file name of the photo, check the contents, and if there are any defects, click the "Cancel" button and try again. If there is no problem in the content, click the "Register" button to transfer the data to the information registration server 11 of the server 1 and store it. This is the end of the user registration procedure (FIG. 3: S4).

Next, the distribution determination process (S3) shown in FIG. 3 will be described. Here, FIG. 6 is a flowchart showing in detail the procedure of the delivery determination process. Here, the information parameter, the interest parameter, the transmission intensity, and the reception sensitivity obtained by the above procedure are read (S31). Then, the fitness is calculated based on these numerical values (S32).

Here, FIG. 7 shows the processing of the fitness calculation (S3
It is a flow chart which shows the procedure of 2) in detail. FIG. 8 is a diagram showing an example of information parameters. FIG. 9 shows initial values of interest parameters. FIG. 10 shows an initial value of transmission intensity, and FIG. 11 shows an initial value of reception sensitivity. Fitness calculation (S
32) first normalizes the information parameters (FIG. 7:
S321) Here, as shown in FIG. 8, the information parameter of "wooden cabinet" is "wooden product / pottery / interior article /
The value of each element of "kitchen supplies / storage goods" is "2,0,
2, 0, 1 ". Hereinafter, this parameter is shown as (2,0,2,0,1). First, when the value of each parameter is squared, it becomes (4,0,4,0,1), and the sum SUM = 9 and its square root SQ.
Determine RT = 3. Next, each parameter (2, 0, 2,
0,1) is divided by this SQRT = 3 to obtain a normalized value (0.667, 0.000, 0.667, 0.000,
0.333).

Next, the interest parameters are normalized (FIG. 7: S322). As shown in FIG. 9, the initial value of the interest parameter of the trader 4 is “3, the value of each element of“ woodenware, pottery, interior goods, kitchen equipment, and storage goods ”.
1,0,1,1 ". Hereinafter, this parameter is shown as (3, 1, 0, 1, 1). First, if the value of each parameter is squared, it becomes (9,1,0,1,1), and the sum SUM = 12 and its square root S
QRT = 3.464102 is calculated. Next, each parameter (3, 1, 0, 1, 1) is converted into SQRT = 3.464.
Divide by 102 to obtain the normalized value (0.866, 0.2
89, 0.000, 0.289, 0.289).

Then, the inner product of each parameter is calculated (FIG. 7: S323). For each element of "wooden products, pottery, interior products, kitchen products, storage products", the value of the information parameter of the normalized wooden cabinet (0.667,
0.000, 0.667, 0.000, 0.333)
And the value of the normalized interest parameter (0.866,
0.289, 0.000, 0.289, 0.289). Then, as shown in FIG. 12, (0.57735,0,0,0,0.09622
The value of 5) is obtained.

Then, the reference value is calculated (FIG. 7: S32).
4). The product of each element of "wooden products, pottery, interior products, kitchen products, storage products" (0.
57735, 0, 0, 0, 0.096225) to obtain the inner product. FIG. 13 is a diagram showing an example of a criterion for determining the availability of distribution. If the transmission intensity and the reception sensitivity are not taken into consideration, a threshold value for determining whether or not the distribution is performed here is determined in advance. Here, assuming that the threshold value is set to 0.65, when the value "0.673575" obtained as the inner product is the reference value, this reference value is compared with the threshold value, and the reference value is larger than the threshold value. If they are equal, it is decided to deliver. In this case, since the reference value 0.673575> the threshold value 0.65, the “delivery” decision is made. In addition, if the reference value <the threshold value, it is determined that “no distribution”.

In this embodiment, the transmission intensity and the reception sensitivity are set as described above. Therefore, in order to calculate the determination value in consideration of the transmission intensity and the reception sensitivity, the following calculation is further performed (FIG. 7: S325). First, in this example, FIG.
It is assumed that the transmission strength is set to "1" as shown in. Further, as shown in FIG. 11, it is assumed that the reception sensitivity is set to "3". And this transmission strength = 1
Then, the product of the reception sensitivity = 3 and the reference value of 0.673575 is obtained. In this way, the determination value = 2.020762 considering the transmission intensity and the reception sensitivity is obtained. With this, the process of calculating the fitness (FIG. 6: S32) is completed (FIG. 7: end).

FIG. 14 is a diagram showing a table for finally determining whether or not distribution is possible. Here, the determination value 2.020762 obtained by the procedure as described above is compared with the predetermined threshold value 2.4, and if the determination value ≧ the threshold value, it is determined to be “delivered”. . In this example, the determination value is 2.020762 <threshold value 2.4, so that the decision "not deliver" is made. Here, the information determined to be “distributed” is stored as a transmission file of the distribution server 13 for each user.

Next, returning to FIG. 6, the description will be continued. A random number generator built in the server 1 generates a random number, and an uncertainty factor is added based on this random number (S33). By adding this uncertainty factor, the degree of fitness is calculated once (S3
Even if the determination value for not delivering is calculated by 2), if an uncertain factor based on a random number or a pseudo random number (simply referred to as a random number in the present application) is added, the predetermined threshold is exceeded ( (S34: YES), and as a result, it becomes a target of distribution (S35).

Processing for adding an uncertain factor (S33)
Then, the information randomly selected from the information other than the information for which the matching degree calculation (S32) is performed and the distribution is determined is processed. The information calculated and selected for the goodness of fit is a result that reflects the interest of the user, but the interest of the actual user is
Information that the user chooses if he / she has no interest, because the new stimuli received up to that point will constantly change it in a fluid manner, and giving new stimuli may create new fields of interest. Even so, I hope that by sending this, new interest will be generated and the possibility of transactions will be created. As described above, the process of adding an uncertain factor (S33) prevents the field of information to be distributed from converging within a certain range and becoming rigid.
There are always opportunities to open up new fields.

The uncertain factor is, for example, that the numerical value is rewritten only when the last digit of the randomized integer is 0,
In the case of 0 to 3, it is possible to control the amount of information to be added at random by controlling the amount of information to be added by changing the setting such as rewriting the numerical value. In this way, the method of uniformly rewriting the interest parameter relating to the information selected by the random number to the highest value (2, 2, 2, 2, 2), or the information parameter is changed to (2, 2,
Rewrite as 2, 2, 2), and further set the distribution strength to "5"
Or by setting the reception sensitivity to “5”, the determination value can be increased. Note that if the method is to adjust the distribution intensity or the reception sensitivity, the influence of the interest parameter and the information parameter remains, and the selection is made in consideration of the interest of the user and the attribute of the product to some extent.

Next, the judgment value to which the uncertainty factor is added is
The information that is compared with the threshold value (S34) and the judgment value exceeds the threshold value (S34: YES) is calculated first in the conformance degree (S3).
In addition to the information whose distribution has been decided by the judgment value calculated in 2), the information whose distribution has been decided here is additionally stored in the transmission file of the distribution server 13 (S35). If the judgment value does not exceed the threshold value (S34: N
O), the distribution process is not performed (S36). When the processing of one information is completed, the processing of other unprocessed information is performed (S3
7: YES → S31), when there is no information to be processed (S37: NO), the distribution determination process (FIG. 3: S3) is ended (END), and the process proceeds to the distribution process (FIG. 3: S4).

In the distribution process (FIG. 3: S4), the information accumulated in the transmission file of the distribution server 13 is collected for each user, and the trader 4 who is the user via the Internet 2
Deliver by email to.

Subsequently, the reply of the information evaluation, which is the evaluation of the user with respect to the distributed information, is received (FIG. 3: S5).
Here, FIG. 15 is a diagram showing the information list screen P3 delivered to the user by e-mail. Here, “mail to recipient 1”, “link to information A”, “link to information B”, “link to information C” ... Are displayed. A hyperlink is set in the character portion of this display so that the displayed information can be accessed. When selected from “Link to information A”, “Link to information B”, “Link to information C” ... and clicked, the URL of the product information detail screen P4 related to this product is the WEB server 14 of the server 1.
Then, the server 1 displays the product information detail screen P4 of the selected product on the terminal of the trader 4.

FIG. 16 is a view showing the product information detail screen P4 displayed on the terminal of the trader 4. Here, "rattan basket (round)" as "product name", "handmade varnish finish" as "outline", and "40 cm" as "specification"
(Diameter) x 60 cm (height) "and" price "as" 1.
50 US $ (FOB) "," 500 "as" lot "
As described above, the content of the product presented by the manufacturer and the photograph of the product are displayed.

At the same time, the product information detail screen P4 displays "Cool !!" and "Goo" when "interested".
d! , "Standard", "NoGood" 4
Evaluation of the stage and "Out of when no interest"
"interest" is displayed. The user stores the evaluation by clicking the icon displayed corresponding to the evaluation. The user can later display a list of emails saved by the user. At this time, the importance of the emails saved by the user is displayed for each email, and the user sorts the emails based on the user's own evaluation.・ It can be utilized. On the other hand, to the server 1, the selection of either "interested" or "not interested" for the product of the user is transmitted as the information evaluation.

The server 1 stores the returned data in the evaluation storage server 17. This corresponds to the step of receiving the information evaluation of the present invention. In addition, the information evaluation is not limited to the two-level evaluation such as "interested" and "not interested" as shown here, or the three-level evaluation that further subdivides the case of "interested". Good. The evaluation storage server 17 has a reaction pool 17a set for each user.
There is. Here, FIG. 17 is a diagram conceptually showing the stored contents of the reaction pool 17a. As shown in FIG. 17, the reaction pool 17a corresponds to “interesting parameters”, “wood products”,
"Pottery", "Interior supplies", "Kitchen supplies",
Numerical values stored in the item "storage goods" are stored. The initial values of this numerical value are all set to 0.

Here, it is assumed that the trader 4 selects one of the "interested" evaluations for the transmitted "rattan basket (round shape)" and returns it as an information evaluation. Then, as shown in FIG. 18, the information parameter of the “rattan basket (round type)” is a value obtained by normalizing (2,0,2,0,1) (0.6667,0.0000,0). .666
7, 0.0000, 0.3333) is the reaction pool 17
stored in a. At this stage, the value of the interest parameter is not changed yet. Next, it is assumed that the information regarding "wooden cabinet" is transmitted. The information parameter of this wooden cabinet is (2,0,0,2,0). Then, it is assumed that one of the evaluations of "interested" is again selected and returned by the terminal of the trader 4. Then, the information parameters are normalized (0.6667, 0, 0.3333, 0,
0.6667) is added to the reaction pool 17a and accumulated. The accumulated value of the reaction pool 17a is (1.3333, 0.0000, 1.000) as shown in FIG.
0, 0.0000, 1.0000).

Here, the threshold for executing feedback with the value of the reaction pool as the parameter of interest is called the "reaction pool threshold". It is assumed that the reaction pool threshold is set to "1". Further, the ratio of the numerical value in the reaction pool 17a rounded down to the nearest whole number and fed back to the user's interest parameter is called a "reaction pool addition coefficient". It is assumed that this reaction pool addition coefficient is set to "1". Here, the value of the reaction pool 17a accumulated as described above is (1.33333, 0, 1, 0,
In the case of 1), the numerical value accumulated in the reaction pool 17a exceeds the reaction pool threshold value “1”, and thus feedback to the interest parameter is executed. If this reaction pool threshold value is set large, the frequency of feedback becomes small, and if it is set small, the frequency of feedback becomes large. And the value that is fed back is
Accumulated in the reaction pool 17a (1.33333,0,
(1,0,1) is converted into an integer (1,0,1,0,1),
Numeric value (1, 0, 1,
0, 1). By setting this reaction pool addition coefficient large, the feedback ratio becomes large, and conversely, by setting it small, the feedback ratio can be made small.

By executing this feedback,
Until now, the user's interest parameters have been (3, 1, 0, 1,
If it is 1), as shown in FIG. 20, after the interest parameter is corrected, the fed back values are added to be (4, 1, 1, 1, 2). The numerical value less than 1 accumulated in the reaction pool is directly accumulated in the reaction pool 17a without being fed back. By modifying the interest parameter in this way, the determination value in the above-described fitness calculation changes, and the ratio of “delivery” determination increases.

On the other hand, a case where "no interest" is selected for the distributed information will be described. Here, it is assumed that the value of the reaction pool 17a is (0,0,0,0,0). From this state, assume that "Shamoji" has been distributed. The information parameter of "smoked rice" is (2,0,0,2,0). When the user selects "No interest" for this information and replies, a value obtained by normalizing this information parameter is subtracted from the reaction pool 17a, and (-0.7071,
0.0000, 0.0000, -0.7071, 0.0
000) is accumulated. Furthermore, it is assumed that the information about the "wooden cabinet" is distributed and the information parameter of this wooden cabinet is (2,0,1,0,2). Here, when the user selects “not interested” again and distributes it, from the reaction pool 17a, the normalized numerical value (−0.6667, 0.0000, −0.
3333, 0.0000, -0.6667) is subtracted. Then, in the reaction pool 17a, (-1.37
38,0, -0.3333, -0.7071, -0.6
The numerical value 667) is accumulated.

Since the reaction pool threshold value "1" is determined by the magnitude of the absolute value, the numerical value accumulated in the reaction pool 17a is fed back to the interesting parameter. Here, the initial values of the interest parameters are (3, 1, 0,
(1, 1), by executing this feedback, (-1, 0, 0, 0, 0) is added, and the interest parameter is corrected to (2, 1, 0, 1, 1). In this case, the numerical value less than 1 accumulated in the reaction pool 17a is accumulated as it is.

As described above, the reaction pool threshold value is set, and when the numerical value accumulated in the reaction pool 17a exceeds this threshold value, the interest parameter is corrected so that the interest parameter can be obtained by only one reply. When the reaction pool 17a does not suddenly change significantly and a numerical value equal to or larger than the threshold value is accumulated, it is possible to gradually correct the user's interest as a change.

Although the reaction pool threshold is set to "1", the sensitivity of correction of the interest parameter with respect to the user's information evaluation can be adjusted by changing this threshold. Although the reaction pool addition coefficient is set to "1" here, the value of the feedback of the information evaluation for the current interest parameter can be adjusted by changing this value.

In this embodiment, the numerical value accumulated in the reaction pool is configured to gradually decrease with the passage of time. Here, the reaction pool damping coefficient is set to "0.9", and the reaction pool damping coefficient is multiplied each time one day has elapsed, and the absolute value is gradually decreased at a constant rate. Specifically, it is calculated as “Today's cumulative value = cumulative value up to yesterday × 0.9 + value to be added today”. Alternatively, the constant value is subtracted from the absolute value to gradually decrease the absolute value. If you happened to be interested in something and replied to it, but it was temporary, and you didn't appear to be interested in similar information after that, reaction pool 1
It is meaningless to leave a numerical value regarding this information evaluation in 7a and feed it back. Therefore, as long as the replies with the same tendency do not continue, the effect of temporary information evaluation is reduced with the passage of time.

Note that, similarly to the present embodiment, a numerical value may be accumulated in the reaction pool 17a based on the information evaluation returned from the user, and this may be fed back at regular intervals. With this configuration, it is possible to prevent the long-term information evaluation from being fed back to the interest parameter.

In the present embodiment, the reaction pool threshold is used to feed back the interest parameter when the value of the reaction pool 17a exceeds a certain value. However, for example, each time there is a reply from the user. The contents may be sequentially fed back to the interest parameter. With this configuration, information evaluation from the user can be reflected in the interest parameter in real time.

In the present embodiment, for the information delivered to the user, the time from the delivery to the user to the reply from the user is measured, and this time is shorter than the predetermined reference time. When the time is longer, the numerical value accumulated in the reaction pool is increased, and conversely, when the time is longer than the predetermined time, the numerical value accumulated in the reaction pool is decreased. It can be presumed that a user who responds quickly to the information evaluation from the user after delivering the information in this way has a particularly strong interest in the information. Increase the degree of correction to.

Further, in the present embodiment, as shown in FIG. 15, a plurality of pieces of information are simultaneously delivered to the same user at one time. When only a part of the plurality of pieces of information delivered at the same time is referred to and the information evaluation is returned, it is considered that the referred information is of high interest and the numerical value accumulated in the reaction pool 17a is increased. On the other hand, among the information distributed at the same time, information that has not been referenced is considered to have relatively low interest, and the value accumulated in the reaction pool 17a may be reduced,
Or reduce it. With such a configuration, it is possible to set the interest parameter that more tends to the interest of the user.

In the present embodiment, the interest parameters are modified by combining the above-mentioned judgments. Figure 2 here
FIG. 1 is a flowchart showing in detail the procedure of the data update process (S6) in FIG. First, when the information evaluation reply is received from the trader 4, it is determined whether the information is referred to within a certain period after the information is provided (S61), and when the information is not referred to within the certain period (S61: NO). Further, it is determined whether or not one or more pieces of information provided at the same time are referred to (S62).

Here, if one or more pieces of information provided at the same time are not referred to (S62: NO), the interest parameter is not corrected because the trader's intention is not known (S63). On the other hand, the information was not referenced (S
61: NO), but if one or more pieces of information provided at the same time are referred to (S62: YES), the trader has an opportunity to refer to this information but does not refer to the information, and the Decrease the value (S64).

On the other hand, when the information is referred to within a certain period after the information is provided (S61: YES), it is determined whether any evaluation of "interested" is selected or "not interested" is evaluated. If it is evaluated (S65) and "no interest" is evaluated (S65: NO), the interest parameter is lowered (S66). If any of the evaluations of "interested" is selected (S65: YES), the interest parameter is increased (S67). When the correction of the parameter of interest is completed, the data updating process (S6) of FIG. 3 is completed (END). This is performed for each reply of the information evaluation, and the interest parameter is corrected. If there is data to be processed (S7: NO), S1
Through S6 are repeated, and if the process is completed (S
(7: YES), all processing ends (END).

As described above, the present embodiment has been described by exemplifying the case where the information transmitted from the arbitrary maker 3 is distributed to the trader 4 which is the user. However, the arbitrary trader 4 similarly acquires the information from the maker. 5 can be redistributed to the buyer 5, and information sent by the trader 4 itself can be delivered to the buyer 5. In this case, from the trader 4 to the manufacturer 3,
Alternatively, the buyer 5 orders the trader 4 with the product via the server 1, but this is not described here.

(Effects of the Embodiment) According to the server 1 of the above embodiment, the following effects can be obtained. By taking the interest of the trader 4 receiving the information distribution as the interest parameter and dividing the distributed information into the information parameters, the tendency of the user's interest can be grasped more accurately, and the attribute of any information distributed is also more accurate. By analyzing the above, information having a tendency closer to the interest of the trader 4 can be distributed. Therefore, the trader 4 can receive only the information more suitable to his / her interest, and the manufacturer 3 who distributes the information can more efficiently and efficiently distribute the necessary and sufficient information to the trader 4. There is an effect.

Further, since the inner product is obtained by normalizing each element of the interest parameter and the information parameter and the degree of coincidence of each element of the user's interest and information attribute is judged from this value, more accurate judgment can be made. There is an effect.

Further, there is an effect that the quantity of distribution can be controlled by the transmission intensity and the reception sensitivity.

Further, by randomly adding information that is not delivered because the judgment value is below the threshold value as it is, it is possible to deliver information in a new range of interest and cultivate a new interest of the user. This has the effect of preventing the contents of distribution from being fixed.

By adjusting the transmission strength and the reception strength of the randomly selected information, it is possible to select the information that is relatively distributed to the user's interest and the information attribute, and is completely irrelevant. This has the effect of reducing the number of large emails delivered.

There is an effect that the reaction of the user can be fed back to the interest parameter by the information evaluation returned by the user to respond to the change in the interest of the user.

Furthermore, regarding the information distributed to the user,
By modifying the interest parameter in accordance with the length of time from the delivery to the reception of the information evaluation from the user, it is possible to deliver the content more in line with the user's interest.

Also, by providing a difference between the information for which the information evaluation has been returned and the information parameter for the information for which the information evaluation has not been returned, there is an effect that it is possible to perform distribution that more matches the interest of the user.

Further, if the information evaluation is sequentially fed back, there is an effect that the information evaluation from the user can be reflected in the interest parameter in real time.

There is an effect that the interest parameter can be corrected at the stage when the change has accumulated, without the needless correction of the interest parameter due to the temporary change of the information evaluation.

Furthermore, there is an effect that the interest parameter can be corrected by a continuous change without making unnecessary correction by the temporary change of the information evaluation of the interest parameter.

In the present application, “direct mail” includes electronic mail on the Internet,
Included in various communication systems where the user can return some response to the distribution server. ○ “Computer” means a mobile phone, a PDA (Personal), whether stationary or mobile.
Digital Assistant), STB (Se
t-Top Box), regardless of its name, includes a CPU and is capable of communicating with a network. It should be noted that the server 1 is composed of a plurality of computers in this embodiment, but may be composed of one computer or may be equipped with another computer.

(Other Examples) Next, the technical ideas that can be understood from the above-described embodiment and other examples will be added below along with their effects. ○ A computer connected to a network, which divides the field of interest of the user into predetermined elements and stores the interest parameter storage means for storing the interest parameters as a combination of the parameters of each element, and the attribute of the information to be distributed. Information parameter storage means for dividing the same element as the interest parameter and storing it as an information parameter consisting of a combination of parameters of each element, and whether or not the information can be distributed by performing calculation based on the interest parameter and the information parameter. An information distribution apparatus in a network, comprising: a distribution availability determining unit that determines the distribution availability; and a distribution unit that distributes the information, the distribution of which is determined by the distribution availability determination means, via the network. A computer connected to a network divides the fields of interest of the user into predetermined elements, and stores them as an interest parameter storage unit that stores the interest parameters as a combination of parameters of each element, and the attributes of the information to be distributed. Information parameter storage means for dividing the same element as the parameter and storing it as an information parameter composed of a combination of parameters of each element, and determining whether or not to distribute the information by performing calculation based on the interest parameter and the information parameter An information distribution program in a network, which functions as a distribution permission determination function and a distribution unit that distributes the information determined to be distributed by the distribution permission determination unit via the network.

[0100]

As is apparent from the above description, the method of distributing information in the network of the present invention analyzes the user's interest in a more actual form, and direct mail relating to arbitrary information can be delivered to the appropriate user. There is an effect that it can be delivered to necessary and sufficient.

[Brief description of drawings]

FIG. 1 is a conceptual diagram showing an entire network to which information is distributed.

FIG. 2 is a conceptual diagram showing a system configuration of a server 1.

FIG. 3 is a flowchart showing an outline of a flow of information distribution in the server 1.

FIG. 4 shows an example of a user registration screen P1 displayed on the terminal of the trader 4.

FIG. 5: Information registration screen P displayed on the terminal of maker 3
An example of 2.

FIG. 6 is a flowchart showing in detail the procedure of distribution determination processing.

FIG. 7 is a flowchart showing in detail the procedure of the process of calculating the goodness of fit (S32).

FIG. 8 is a diagram showing an example of information parameters.

FIG. 9 is a diagram showing initial values of interest parameters.

FIG. 10 is a diagram showing an initial value of transmission intensity.

FIG. 11 is a diagram showing an initial value of reception sensitivity.

FIG. 12 is a diagram showing inner products when the normalized information parameters of the wooden cabinet and the normalized interest parameters are regarded as vectors.

FIG. 13 is a diagram showing an example of a criterion for determining whether or not distribution is possible.

FIG. 14 is a diagram showing a table for finally determining whether or not distribution is possible.

FIG. 15 is a diagram showing an information list screen P3 distributed to a user by e-mail.

FIG. 16 is a view showing a product information detail screen P4 displayed on the terminal of the trader 4.

FIG. 17 is a diagram conceptually showing stored contents of a reaction pool 17a.

FIG. 18 is a diagram showing normalized values of information parameters of “rattan basket (round type)” stored in the reaction pool 17a.

FIG. 19 is a diagram showing accumulated values of the reaction pool 17a.

FIG. 20 is a diagram showing interest parameters after being corrected by performing feedback from the state of FIG. 9;

21 is a flowchart showing in detail the procedure of the data update process (S6) in FIG.

[Explanation of symbols]

1 ... Server 1, 2 ... Internet, 11 ... Information registration server, 12 ... Recipient DB, 13 ... Distribution server, 14 ... W
EB server, 15 ... Binary data storage server, 17 ...
Evaluation storage server

Claims (15)

[Claims] 1. A step of storing an interest parameter in which a field connected to a network is divided into predetermined elements in a computer connected to a network and stored as an interest parameter consisting of a combination of parameters of each element, and information to be distributed. Of the attribute of the interest parameter is divided into the same element, the information parameter storing step of storing as an information parameter consisting of a combination of parameters of each element, and information by performing a calculation based on the interest parameter and the information parameter Information distribution in a network, characterized by executing a distribution permission / inhibition determination step of determining whether or not the distribution is possible, and a distribution step of distributing the information determined to be distributed in the distribution availability determination step via the network. Method. 2. In the step of determining whether or not the distribution is possible,
The inner product of the parameters of the corresponding elements of the interest parameter and the information parameter is calculated, this value is compared with a predetermined threshold as a reference, and whether the distribution is possible or not is determined based on the result. An information distribution method in the network described in 1. 3. In the step of determining whether or not the distribution is possible,
The information distribution method in a network according to claim 2, further comprising a step of setting a distribution strength for uniformly modifying each parameter of the information parameters with the same magnification. 4. The method according to claim 1, further comprising a step of setting a receiving sensitivity for uniformly modifying each of the parameters of interest at the same magnification according to a user's request.
An information distribution method in the network according to claim 3. 5. The information is extracted based on a predetermined extraction method from the information which is determined not to be distributed in the distribution permission / inhibition decision step, and is distributed in the distribution step. Item 5. An information distribution method in the network according to any one of Items 4. 6. The predetermined extraction method is performed by making a predetermined change to any one or more of the information parameter, the interest parameter, the distribution intensity, and the reception sensitivity, and then comparing with the threshold value. The information distribution method in a network according to claim 5, wherein 7. The information distribution method in a network according to claim 5, wherein the predetermined extraction method includes a step of extracting based on a random number. 8. An information evaluation receiving step of receiving from the user an information evaluation, which is an evaluation given to the information by the user who received the information, and based on the information evaluation received by the information evaluation receiving step. The information distribution method in a network according to any one of claims 1 to 7, further comprising a first interest parameter modification step of modifying an interest parameter corresponding to the information parameter. 9. The second parameter for modifying the interest parameter corresponding to the information parameter based on the length of time from the time of distribution to the time when the information evaluation from the user is returned, for the information distributed to the user. 8. The method according to claim 1, further comprising a step of modifying an interest parameter.
An information distribution method in the network according to any one of 1. 10. A method of correcting an interest parameter corresponding to an information parameter of information for which no information evaluation is returned, when an information evaluation of only a part of the plurality of information simultaneously delivered to the user is returned. The information distribution method in a network according to any one of claims 1 to 7, further comprising the step 3 of modifying an interest parameter. 11. The information distribution method in a network according to claim 8, wherein the modification of the interest parameter is sequentially performed each time a user returns an information evaluation. 12. The modification of the interest parameter comprises a reaction pool for temporarily accumulating data based on a reply of information evaluation from a user, and a predetermined numerical value accumulated in the reaction pool is equal to or larger than a predetermined threshold value. 11. The method according to any one of claims 8 to 10, further comprising the step of modifying the parameter of interest when the information becomes negative. 13. The information distribution method according to claim 12, wherein all or part of the numerical values accumulated in the reaction pool are gradually reduced or discarded as time passes. 14. The modification of the interest parameter is provided with a reaction pool which temporarily stores data based on a reply of information evaluation from a user, and a predetermined numerical value accumulated in the reaction pool at a constant time interval. The method according to any one of claims 8 to 10, further comprising the step of modifying the interest parameter based on the information distribution method. 15. An information evaluation receiving step of receiving, from the user, an information evaluation, which is an evaluation given to the information by the user who has received the information, distributed information, and the distributed information related to the distributed information. 15. The received information list list display step of displaying a list on the user's terminal in association with the information evaluation received in the information evaluation reception step, according to any one of claims 1 to 14. Information distribution method in the Internet.