JP4574292B2 - Advertisement providing method, advertisement providing apparatus, and computer program - Google Patents

Description

  The present invention relates to an information delivery method for delivering information (items) on media such as the Internet, digital interactive television, and other information home appliances, and an advertisement providing method in interactive media.

  In recent years, the use of the Internet and the like has been remarkable, but the user can only browse information such as accessed Web pages or obtain a list of search results. For example, if you search for books on a book sales site, you can get a list of books that match your search criteria, but you can find other books and other information (related to users) that are related (correlated) to them. Cannot be easily obtained.

  Also, in the marketing information analysis, pre-processing such as an evaluation value for each item input by the customer or genre classification for the item is assumed. In addition, complicated conditions for correlation detection need to be set.

  On the other hand, interactive media, such as the Internet, which can immediately return a reaction to an advertisement, has attracted attention because it can more easily acquire a user. In such media advertisements, it is important not only how many users can see the advertisement, but how many users have issued a request for acquiring additional information.

  There are mainly the following methods for displaying advertisements in such interactive media.

(1) Random method
(2) How to put out according to a certain content
(3) Method using user's static attribute information
(4) How to ask the user to explicitly enter what they are interested in
(1) and (2) are general methods used in conventional mass media. Since newspaper advertisements are widely read, they fall under (1). Moreover, since a television advertisement etc. are provided for every specific program, it corresponds to (2). Even in the Internet WWW service, there is a method of advertising only for specific contents.

  In (1) and (2), the users who actually refer to the advertisement are not clear, so the methods (3) and (4) are mainly used in interactive media in order to increase requests for additional information. Is used. However, (3) and (4) have the following problems.

  i) The user must enter information such as user attributes one by one. It is difficult for the user to spend such effort.

  ii) Since the information entered when there is a user is used, the user information is fixed or obsolete.

  iii) I don't know if the user enters correct information.

  Conventionally, in the notice distribution apparatus via the Internet or a network, an advertisement of an item of the same or similar genre to an item simply purchased by a customer in the past is delivered to the customer, and the same genre as the advertisement target item To be distributed to customers who have access to information related to the item (Double Click banner advertising distribution technology), the past purchase history is subjected to analysis such as correlation analysis, the hobby and age group of the customer who purchased the item to be advertised, The distribution of attributes such as gender and occupation is extracted and distributed to customers having an attribute accounting for a large proportion of the attribute distribution.

  However, the conventional advertisement distribution apparatus has the following problems.

  -Pre-registration of attribute information such as hobby, age, occupation, etc. is necessary for the customer, and the pre-registration of the attribute becomes a barrier, and the number of customers does not increase.

  Unless the attribute registration is updated, even if there is a change in the customer attribute, the change in the attribute is not reflected.

  ・ Registration of attribute information such as a genre (category) is necessary for an item, and management is troublesome. Here, if the genre is only roughly classified, the management troublesomeness is reduced, but for example, when Hollywood movies, French movies, and Japanese movies are grouped into the same `` genre '' genre, Distribution of advertisements for French movies to people who have a hobby for “movies” is not effective for those who like Hollywood movies, so the accuracy of narrowing down the distribution destinations of advertisements is worse.

  -Since the distribution destination of the advertisement is restricted by the category and the genre, the advertisement cannot be distributed to a customer (potential customer) who has a correlation and exceeds the genre and has a high possibility of being interested.

  It is an object of the present invention to automatically extract users who are likely to be interested and users who are likely to be interested in the advertisement posting in interactive media without the need for users to input information one by one. The challenge is to provide the user with the optimal advertisement.

  Moreover, this invention makes it a subject to provide an advertisement to a potential customer so that an advertisement may be delivered only to the customer with the high probability of purchasing an item, without registering the attribute of a customer and the attribute of an item.

In order to solve the above-described problem, the present invention provides a client in which an advertisement providing apparatus is a user who is accessing an access history of user access to information provided by an information provider and target information specified in advance. both steps and, the additional information virtual user and the object information client an advertisement information to be managed by storing the advertisement information associated with the access possible additional information in the database provided the history that has been accessed, as said access history in association with the unique identifier and said object information and said additional information identifying the user of the virtual, and the step of further stored in the database, the advertisement information selection means, said based on the access history, associated use of the user's virtual there is history of accessing the target information other than the client As the extraction, detects a unique identifier associated with the subject information, as the correlation information being accessed by the associated user based on the access history, the additional information associated with the identifier of said intrinsic And providing the advertisement in response to a request from the client by delivering the advertisement information managed in association with the extracted additional information to the client. I will provide a.

In the present invention, potential target information generating means, and detecting the other related users who access the target information other than the user of the virtual and the client based on the access history, potential target information generation history that means, extracting potential target information accessed by said other related user based on the access history, another virtual user has access to both the additional information and the target information and as said access history, and the step of storing in a database, the advertisement information selection means, to another client that accesses the potential target information, based on the access history, the potential target client except the another detecting the another virtual user as another related user who has accessed the information, said further based on said access history Extracting the additional information as correlation information accessed by a related user, and delivering the advertisement information managed in association with the extracted additional information to the other client. Features.

  In the present invention, the providing step provides the advertisement information in a form linked to the additional information.

  In the present invention, the managing step manages the advertisement information designated by the advertiser, the additional information, and the target information.

In the present invention, advertisement providing count management means, comprising the steps of Ru is the actual advertisement providing operation by storing in the database by aggregating the additional information access count and the advertisement information delivery count, the advertising fee calculating means, the advertisement information a step of charging advertisers advertising fees based on at least one of the delivery number and the additional information access count, and further comprising a.

In the present invention, the advertisement fee calculation means, so that the payment of prior advertisement fee to an advertiser actual advertisement providing operation, minimum advertisement information delivery count and minimum additional information that is guaranteed is stored in the database The advertiser is charged with an advertising fee for at least one of the access counts , and the advertisement provision count management means totals the advertisement information delivery count and the additional information access count by the actual advertisement provision operation and stores them in the database. And the advertisement provision frequency management means until the advertisement information delivery count reaches the minimum advertisement information delivery count, or until the additional information access count reaches the minimum additional information access count, or The advertisement information delivery count reaches the minimum advertisement information delivery count, and the additional information access count is the minimum additional information access count. Until it reaches the number of times, characterized by further comprising the steps of extending the advertisement posting period for the actual advertisement providing operation, a.

In the present invention, the advertisement fee calculation means, so that the payment of prior advertisement fee to an advertiser actual advertisement providing operation, minimum advertisement information delivery count and minimum additional information that is guaranteed is stored in the database The advertiser is charged with an advertising fee for at least one of the access counts , and the advertisement provision count management means counts the advertisement information delivery count and the additional information access count by the actual advertisement provision operation during a predetermined advertisement posting period. a step of Ru is stored in the database, advertisement fee calculation means, the advertisement information delivery count the minimum advertisement information of the difference between the difference and the minimum additional information access count of the additional information access count and delivery count at least one of And a step of refunding at least a part of the advertising fee corresponding to either of them.

Further, the present invention provides an access history of access by a user to information provided by an information provider, and advertisement information to be provided to a client who is a user who is accessing target information specified in advance . means clients managed by storing the advertisement information associated with the access possible additional information in the database, a history of the virtual user has access to both the additional information and the target information, the virtual It means Ru is further stored in the database as the access history by associating a unique identifier with said object information and said additional information identifying the user of, based on the access history, history of accessing the target information other than the client as a related user virtual user that detects a unique identifier associated with the target information, As the correlation information being accessed by the associated user based on the serial access history, and extracts the additional information associated with the identifier of said intrinsic, extracted ads that are managed in association with the said additional information There is provided an advertisement providing apparatus characterized by having means for delivering an advertisement in response to a request from a client by delivering information to the client.

Furthermore, the present invention includes an access history of the access by the user for the information provided by the information provider, click an information to be provided to the client is a user who is accessing the pre-specified target information client There the steps of Ru is managed by storing the advertisement information associated with the access possible additional information in the database, a history of the virtual user has access to both the additional information and the target information, the virtual as said access history in association with the unique identifier and said object information and said additional information identifying the user of the steps of Ru is further stored in the database, based on the access history, and access the target information other than the client the virtual user that there is a history as related users, the unique identifier associated with the target information Out, wherein the correlation information being accessed by the associated user based on the access history, and extracts the additional information associated with the identifier of said intrinsic, are managed in association with the extracted the additional information by delivering the advertisement information are in the client, a computer program for advertisement providing characterized by having a step of providing advertisement in response to a request from the client, the.

  According to the present invention, in advertising display on interactive media, a user who is likely to be interested and a user who is likely to be interested are discovered, and an appropriate advertisement is based on the latest access log of the user. Information can be selected and provided. Thereby, in the advertising fee system based on the number of advertisement references and the number of additional information acquisition, even if the number of advertisement references is the same, the number of additional information acquisition increases, and there is an advantage of obtaining more income.

  According to the present invention, if a plurality of pieces of target information are set, a user who refers to a plurality of specific contents can be targeted rather than a user who refers to only one specific content. In addition to target information, the target can be expanded by determining the target users based on latent target information.

  Furthermore, according to the present invention, for a target item to which an advertisement is to be distributed, based on the access history only (pre-registration of the attribute of the customer and the item is unnecessary), a customer who is highly likely to browse / purchase It is possible to deliver advertisements to the network. For this reason, the number of browsing / purchasing can be increased as a result.

  Hereinafter, the first embodiment will be described in detail with reference to FIGS.

  Here, information that can be browsed, held in a basket, etc., purchased by a client via a browser, information such as a product, an advertisement, and a Web page title is referred to as an item. An item can be uniquely identified by an item identifier (ID) such as a Web URL or a product code.

  FIG. 1 is a system configuration diagram of this embodiment. In the figure, 100 is a client, 110 is an information delivery server (item delivery server), 120 is an information browsing server (item browsing server), and 130 is a network such as the Internet connecting them. In general, a large number of clients 100 exist on the network 130.

  The client 100 is a Web-accessible device such as a personal computer, a digital TV, or a mobile phone, and has a browser 105. The browser 105 is information access means such as a Web browser for searching for items, purchasing items, and receiving item recommendations. The client 100 has a meaning as an input unit for performing a contracting action for correlated item delivery to the information delivery server 110 and as an output unit for displaying a list of correlated items. As long as there is, it is not limited to a web browser.

  The information browsing server (item browsing server) 120 is a search server or an EC server that can browse and purchase products and information via a network, and holds an item browsing / purchasing history such as browsing and purchasing items. The item browsing / purchasing history is held in a format including information such as “user identifier, item ID, browsing date”, for example, and there is a means for using the access history of the Web server and a unique history holding means. . The user identifier is an identifier for uniquely identifying a user who uses the information browsing server 120 globally, and is normally assigned to the terminal of the user by the information browsing server.

  Note that there may be a plurality of information browsing servers 120 on the network 130 (generally this is more common), but in that case, items representing the same information and the same product must have the same item ID, In addition, it is necessary to be able to confirm that the user identifier is the same user even across the Web servers.

  For this purpose, although omitted in FIG. 1, the item browsing / purchasing history of each information browsing server is collected in a log management server through a network or the like so that the whole can be managed collectively. User identifiers are also managed in a unified manner. Alternatively, the user identifier is assigned by the information browsing server, but when a unified identifier is assigned by a plurality of information browsing servers, a unique identifier is assigned by querying the log management server.

  The information delivery server (item delivery server) 110 periodically acquires and holds item browsing / purchase history from the information browsing server 120, and uses the stored item browsing / purchase history to deliver correlated items in advance. Delivery of correlated items to the registered client 100.

  The information delivery server 110 includes a Web page 111, a server engine 112, an item correlation detection unit 113, a browsing / purchase history database 114, a client management unit 115, a base item storage unit 116, a correlation item storage unit 117, and a charge management unit 118. Composed.

  The server engine 112 receives an HTTP request for accessing the Web page 111 identified by the URL, and provides the Web page 111 to the client 100. Here, it is assumed that the HTTP request can indicate that the client 100 has performed a registration operation for requesting delivery of a correlated item, and a list of correlated items can be displayed on the client 100.

  The inter-item correlation detection unit 113 periodically performs correlation item detection processing. The correlation item detection method can use a statistical method such as class discrimination, a data mining method using a neural network, a genetic algorithm, demographic clustering, or the like. In the present embodiment, information is collected by social filtering (SF). A method for detecting information correlation (similarity) is used. Social filtering will be described later.

  The browsing / purchase history database 114 holds an item browsing / purchase history that is periodically acquired from the information browsing server 120. The item browsing / purchasing history is held in a format including information “user identifier, item ID, browsing date”.

  The client management unit 115 manages the delivery destination of the correlation item. When the client 100 registers for correlation item delivery for the first time, a client identifier for uniquely identifying the client is assigned globally, and the client identifier is managed together with the keyword received from the client. When accessing 110, the client is identified.

  The base item holding unit 116 holds a list of items from which correlation items are detected together with a client identifier.

  The correlation item holding unit 117 holds a list of the latest correlation items obtained by the detection process of the inter-item correlation detection unit 113 together with the client identifier.

  The charge management unit 118 manages the amount charged for correlated item delivery. It manages a metered charge value according to the fixed amount or amount of information delivered together with the client identifier.

  The operation of this system is broadly divided into access history acquisition, correlation item delivery destination client registration, correlation item delivery registration, correlation item detection, correlation item delivery and accounting, and accounting information aggregation. Hereinafter, the operation will be described in this order.

[Acquire access history]
FIG. 2 shows an access history acquisition processing flow. The information browsing server 120 periodically sends the browsing / purchasing history to the information delivery server 110 (step 201). The information delivery server 110 holds the browsing / purchase history received from the information browsing server 120 in the browsing / purchase history database 114 (step 202). The browsing / purchasing history in the browsing / purchasing history database 114 is deleted from, for example, those after a predetermined period.

[Client Registration]
FIG. 3 shows a processing flow of client registration. This is a process of assigning a client identifier (user identifier) to a correlation item delivery destination client, and needs to be performed prior to the next correlation item delivery registration.

  The client 100 uses the browser 105 to access the information delivery server 110 via the network 130 such as the Internet (step 301), and sends a keyword to the information delivery server 110 together with a client identifier registration request (step 302). The information delivery server 110 generates a globally unique user identifier as a client identifier (step 303), and registers the generated client identifier (user identifier) and the keyword received from the client 100 in the client management unit 115. (Step 304). The client identifier and keyword registered in the client management unit 115 are used to specify the client 100 when accessed from the client 100 thereafter.

  Next, the information delivery server 110 sends the client identifier (user identifier) to the client 100 together with the client identifier registration response (step 305). Upon receiving the client identifier together with the client identifier registration response, the client 100 displays the client identifier (user identifier) (step 306). Thereafter, the client 100 sends the client identifier and the previous keyword sent from the information delivery server 110 to the information delivery server 110, thereby registering the base item and receiving the correlation item.

The client identifier can be held on the client side using a function such as Cookie.

[Register Correlated Item Delivery]
FIG. 4 shows a registration processing flow for correlated item delivery. This is a process of registering a list of base items from the client, a charging type, and the like. The base item list is a correlation detection condition, that is, a list of information (items) from which a correlation item is detected. By registering the base item in the information delivery server 110 in advance, the client 100 can receive delivery of an item group correlated with the base item group described in the base item list from the information delivery server 110. Become.

  The client 100 uses the browser 105 to access the information delivery server 110 via the network 130 such as the Internet (step 401). When accessing the information delivery server 110, the server engine 112 of the information delivery server 110 receives the HTTP request to access the web page (correlation item delivery registration page) identified by the URL, and sends the web page to the client 100. provide.

  The client 100 sends the previously assigned client identifier (user identifier) and keyword together with the client authentication request to the information delivery server 110 (step 402). When the information delivery server 110 receives the client identifier and keyword together with the client authentication request, the information delivery server 110 searches for the client identifier and keyword pair held in the client management unit 115 and confirms that the client is already registered ( In step 403, a client authentication response and a normal authentication message are returned to the client 100 (step 404). If it is not a registered client, the information delivery server 110 returns a client authentication response and a denial message to the client 100 and ends the process.

  When the client 100 receives the client authentication response and the normal authentication message, the client 100 sends the client identifier, the list of base items, and the billing type to the information delivery server 110 together with the correlation item delivery registration request (step 405).

The base item list is a list of items from which a correlation item is detected, and the charge type is, for example, a flat rate system or a pay-as-you-go system. In order to avoid a security risk, the client 100 may always send a keyword together with the correlation item delivery registration request, and the information delivery server 110 may always authenticate the client identifier.

  The information delivery server 110 records the base item list sent from the client 100 together with the client identifier in the base item holding unit 116 (step 406). Next, the information delivery server 110 records the accounting type and accounting information together with the client identifier in the accounting manager 118 (step 407). The billing information is a billing value for a certain period (for example, monthly) when the billing type is a flat rate system, and is a billing value according to the amount of information delivered when the billing type is metered billing. .

[Detection of correlated items]
The information delivery server 110 periodically performs a correlation item detection process (for example, every day). FIG. 5 shows the overall flow of correlation item detection.

  First, a pair of a client identifier and a base item list held in the base item holding unit 117 is extracted (step 501). A client identifier and a list of base item are passed to the inter-item correlation detection unit 113 together with the correlation detection request (step 502). The inter-item correlation detection unit 113 extracts a list of correlated items from the browsing / purchase history database 114 based on the base item list, and holds the correlation item list in the correlation item holding unit 117 together with the client identifier. (Step 503). Thereafter, steps 501 to 503 are repeated until there is no more data held in the base item holding unit 116 (step 504).

  FIG. 6 conceptually shows the correlation item detection. A list of basic items (a list of items from which correlation items are detected) is G1, and a list of users who have viewed G1 other than the users of the base item list is G2. G2 is called a browsing user list. An item list browsed by the browsing user list G2 that does not include G1 is referred to as a correlated item list G3 of G1. That is, G3 is estimated to be of interest to G1 users. Actually, G2 and G3 are given a weight (number of times of browsing), and a predetermined number is selected from those having a higher weight.

  FIG. 7 shows a processing flow of the inter-item browsing detection unit 113. In the inter-item correlation detection unit 113, first, for each item included in the base item list (G1), a browsing user (user) browsing each item is extracted from the browsing / purchase history database 114 and viewed. This is stored as a user list (G2) (step 701). This is repeated for each item in the base item list (step 702). Next, for each user in the browsing user list, count how many times the items included in the base item list are browsed, rearrange them in order of the number of browsing, leave only the top predetermined number, and newly browse This is stored as a user list (step 703).

  Next, for each user in the browsing user list obtained in step 703, the item being browsed by each user is extracted from the browsing / purchase history database 114 and stored as a correlation item list (G3) ( Step 704). This is repeated for each user in the browsing user list (step 705). Next, for each item in the correlated item list, the items included in G1 are removed, and the number of times viewed from each viewer included in the browsing user list is counted, and the items are sorted in descending order of the number of browsing times. Only the upper predetermined number is left, and a new correlated item list is created (step 706).

  This correlated item list is held in the correlated item holding unit 117 together with the client identifier corresponding to the base item list (step 503 in FIG. 5).

In this way, the correlated item holding unit 117 holds the latest correlated item list corresponding to the base item list for each client. When the client changes the base item list, the correlated item list of the correlated item holding unit 117 is also changed correspondingly.

[Correlation Item Delivery and Billing]
FIG. 8 shows a flow of processing of correlation item delivery and billing.

  The client 100 uses the browser 105 to access the information delivery server 110 via the network 130 such as the Internet (step 801). When accessing the information delivery server 110, the server engine 112 of the information delivery server 110 receives an HTTP request to access the web page (correlation item delivery page) identified by the URL, and provides this web page to the client 100. To do.

  The client 100 sends a client identifier and a keyword together with the client authentication request to the information delivery server 110 (step 802). When the information delivery server 110 receives a client identifier and a keyword together with a client authentication request from the client 100, the information delivery server 110 searches for a client identifier and keyword pair held in the client management unit 115, and is an already registered client. (Step 803), a client authentication response and a normal authentication message are returned to the client 100 (step 804). If the client is not a registered client, the information delivery server 110 returns a client authentication response and a denial message to the client 100 and ends the process.

  When the client 100 receives the client authentication response and the normal authentication message, the client 100 sends a client identifier to the information delivery server 110 together with the correlation item delivery request (step 805). In order to avoid a security risk, the client 100 may always send a keyword together with the correlation item delivery request, and the information delivery server 110 may always authenticate the client identifier.

  The information delivery server 110 sends the client identifier from the client 100 to the correlated item holding unit 117 together with the correlated item read request (step 806). The correlated item holding unit 117 reads a list of correlated items that match the client identifier based on the client identifier received together with the correlated item read request (step 807). Next, the information delivery server 110 sends the client identifier and the number of items included in the read list of correlated items together with the correlated item delivery record request to the charge management unit 118 (step 808). The charge management unit 118 records the client identifier, the number of items, and the current date and time (step 809).

  Thereafter, the information delivery server 110 sends a list of correlated items together with the correlated item delivery response to the client 100 that has received the correlated item delivery request (step 810). The client 100 displays the list of correlated items received from the information delivery server 110 on the browser 105 (step 811).

[Aggregation of billing information]
FIG. 9 shows a billing information counting process flow. The information delivery server 110 activates the accounting management unit 118 periodically (for example, every month) and performs accounting processing based on the accounting information.

  The information delivery server 110 sends a billing process request to the billing management unit 118 periodically (for example, every month) (step 901). The charging management unit 118 sequentially reads out the charging type and charging information for each client type (step 902). If the registered billing type is a flat-rate system, it is confirmed that the correlated items are regularly delivered, and then the billing value is integrated as the billed amount (steps 903 and 904). When the registered charging type is a pay-as-you-go system, the charged value times the number of delivery of correlated items is integrated as a billed amount (steps 903 and 905). Thereafter, the charging management unit 118 repeats steps 902 to 905 until the charging processing of all registered clients is completed (step 906).

  FIG. 10 shows a specific processing image of correlated item delivery by this system.

Correlated items are detected based on an access log in which information held by an information delivery server is viewed by a user. For example, an access log obtained by searching for a book on a book sales site and browsing an outline or the like is stored as a log of a WWW server. The stored information is, for example, “access user identifier, information identifier, access date”. In FIG. 10, processing is performed as follows.

  (1) A client as a delivery destination of information (item) registers a correlation detection condition, that is, a base item in advance. The correlation detection condition is a list of information (items) from which a correlation item is detected.

  (2) The information delivery server detects a list of correlated information (items) from the access log based on the information list (base item) given in (1).

  (3) The information delivery server delivers correlated information (correlation item list) to the client for each correlation item delivery request from the client or for each predetermined period.

  (4) The client pays a fixed fee for each amount of information delivered (item amount).

  Although the first embodiment has been described above, the present invention is not limited to this. For example, the following various changes and expansions are possible.

  (1) The network is not limited to the Internet. There are digital interactive TV and information home appliance network.

  (2) Clients include PCs, mobile phones, digital two-way TV sets, information appliances, communicable word processors, communicable game machines, communicable watches, and terminals capable of mobile communication such as PDAs.

  (3) The input / output information medium is not limited to the Web. Using e-mail, a correlated item delivery request, a client identifier, and the like are transmitted as an e-mail document or title. An email address can be used instead of the client identifier. Correlated item lists can also be sent as e-mail documents.

  (4) The correlated item list can be displayed, for example, overlaid on the television screen and over the game display.

  (5) The timing of providing the correlation item list need not be limited to when the client makes a correlation item delivery request. The information delivery server can automatically access the client storage device periodically and write the correlated item list.

  (6) The correlation item detection process does not have to be regular (for example, every day). For example, when a correlation item delivery request is received from a client, a correlation item detection process may be performed, and the detected correlation item may be returned immediately. Thereby, a correlation item holding part can be omitted. The base item may be sent when the client makes a correlation item delivery request. Thereby, a base item holding part can also be omitted.

  (7) There are various variations in the accounting method. For example, at the time of registration of the base item, a minimum guarantee is made with respect to the number of correlated items acquired and a fee is paid in advance. In addition, at the time of registration of the base item, a minimum guarantee is made with respect to the number of correlated items acquired, and a fee is paid in advance. Furthermore, payment is made after the predetermined service period ends based on the total number of correlated items.

  (8) Further, the processing of the inter-item correlation detection unit 113 shown in FIG. 7 can be modified as follows.

  First, as a first modification, a method for extracting trendy items from all item groups will be described.

  First, a base item list (G1) having a total access count equal to or greater than a set value is extracted, and the extracted base item is set as a popular base item list (G1 ').

  For each item included in the popular base item list (G1 ′), a browsing user (user) browsing each item is extracted from the browsing / purchase history database 114. At that time, a preset number of people or users who enter a preset period are extracted in the order of access after a preset date and time. For example, it is assumed that the preset date and time is the oldest date and time in the data, and m people accessing in order after that date and time are taken out. This is the user group who recognized the item first. The user extracted in this way can be regarded as a user who first noticed a popular item, that is, a pioneer in fashion.

  The obtained user group is stored as a browsing user list (G2). This is repeated for each item in the popular base item list.

  Next, from the browsing / purchase history database 114, for each user in the browsing user list, an item being browsed by each user is extracted. At that time, an item accessed at a preset date and time or period is extracted. For example, items accessed in the past n days from the present are extracted. The extracted item is stored as a correlated item list G3. This is repeated for each user in the browsing user list.

  The correlation item list obtained in this way can be regarded as an item that a trendy pioneer is paying attention to, that is, an item predicted to be popular next time.

  Subsequently, as a second modification, a method for extracting trendy items from an item group having a specific attribute and a user group having a specific attribute will be described.

  First, a base item list (G1) having a specific attribute and having a total access count equal to or greater than a set value is extracted. Let the extracted base point item be a popular base point item list (G1 '). The specific attribute is, for example, an item included in a preset genre (eg, cosmetics, clothes, etc.).

  For each item included in the popular base item list (G1 ′), a browsing (user) having a specific attribute is extracted from the browsing / purchase history database 114. Here, the specific attribute is, for example, a user type (for example, female in teens to 30s) based on information obtained from a profile.

  At that time, a preset number of people or users who enter a preset period are extracted in the order of access after a preset date and time. For example, it is assumed that the preset date and time is the oldest date and time in the data, and m people accessing in order after that date and time are taken out. This is the user group who recognized the item first. The user extracted in this way can be regarded as a user who first noticed a popular item, that is, a pioneer of a trend of a specific genre among users of a specific user type.

  The obtained user group is stored as a browsing user list (G2). This is repeated for each item in the base item list.

  Next, from the browsing / purchase history database 114, for each user in the browsing user list, an item being browsed by each user and having a specific attribute is extracted. The specific attribute is the same genre (for example, cosmetics, clothes, etc.) used for obtaining the popular base item list.

  Furthermore, at that time, items accessed from the present date and time or period set in advance are extracted. For example, items accessed in the past n days from the present are extracted. The extracted items are stored as a correlated item list. This is repeated for each user in the browsing user list.

  The correlation item list obtained in this way is the item that the pioneer of the trend of the specific genre is paying attention to, that is, the item predicted to be the next popular in the specific genre among users of the specific user type Can be considered.

  In the first and second modified examples, the correlated items in the extracted correlated item list can be arranged in various orders. For example, as shown in FIG. 7, there is a method of counting how many times each viewer included in the browsing user list has been accessed and rearranging them in order of the number of browsing. In addition, there is a method of counting the number of viewers among the viewers included in the browse user list and arranging them in the order of the number of viewers accessed. In addition, there are some which are rearranged in the order of access date and time.

  Next, a second embodiment related to the advertisement providing method will be described in detail with reference to FIGS.

  In this embodiment, the access history of users of interactive services such as the Internet is used to save the user from having to input each time. When the user refers to the information he is interested in, the information service provider side keeps a reference record. Also, when a specific action is taken, the action record is kept. By collecting this periodically, information at a certain date or time or a reference record collection (access history) for a certain action can be obtained for each user. This reference record collection is used to provide advertisements to users. As a result, the user does not need to input the user attribute or the like again and uses the actually referred information, so that the latest and correct user information can be obtained.

  The technique for providing advertisements to the user using this reference record collection will be described below. Here, an algorithm for calculating the similarity of a plurality of reference record collections is used. This similarity calculation is generally called social filtering (SF).

  FIG. 13 conceptually shows the basic principle of SF information introduction. When the reference information group of the user U0 is G1, a user group other than U0 that refers to G1 is referred to as a related user group G2. A reference information group that is not included in G1 among information groups referred to by related users of G2 is referred to as G3. G3 is regarded as reference information having a high degree of similarity with G1, that is, U0 is estimated as information of interest, and this G3 is provided to U0 as introduction information.

  Here, each related user of G2 is given a value (such as the number of references) indicating a high degree of similarity, and by using that information, each reference information (introduction information) of the reference information group of G3 is used. A value indicating the estimated interest level is assigned. As a result, it is possible to provide a mechanism such as introducing from those with a high degree of estimated interest.

  Also, attributes are set for users and information. Thereby, the related user who becomes G2 can be selected according to the attribute (filtering), and the introduction information that becomes G3 can be selected.

  FIG. 11 is a system configuration diagram of this embodiment. In the figure, 1100 is an advertiser terminal, 1110 is a user terminal, 1120 is an advertisement providing server, 1130 is a log management server, 1140 is an information service providing server, and 1150 is the Internet or other network connecting them. In general, a large number of user terminals 1110 and information service providing servers 1140 exist on the network 1150. A plurality of advertiser terminals 1100, advertisement providing servers 1120, and log management servers 1130 may also exist on the network 1150.

  Here, the advertisement providing server 1120 includes an advertisement information management unit 1121, a user history management unit 1122, a latent target information generation unit 1123, an advertisement information selection unit 1124, an advertisement provision number management unit 1125, an advertisement fee calculation unit 1126, and the like. Is done.

  The advertisement information management unit 1121 manages advertisement information requested by the advertiser to display advertisements.

The user history management unit 1122 manages an access history to each information of the information service providing server 1140, that is, a reference record collection for each user. The user history management unit 1122 further manages an access history of an advertising agent that is a virtual user, which will be described later.

  Based on the user's access history, the latent target information generation unit 1123 generates latent target information having high similarity from target information related to the advertisement information. The advertisement information selection unit 1124 selects optimal advertisement information based on the access history of the user and the access history of the advertisement agent and provides it to the user.

  The advertisement provision count management unit 1125 manages the advertisement reference count and additional information acquisition count of the user. The advertisement fee calculation unit 1126 calculates an advertisement fee to be paid by the advertiser based on the advertisement reference count, additional information acquisition count, and the like.

  First, user access history collection (log collection) will be described. Each user refers to each information of various information service providing servers 1140 using the user terminal 1110, and the information service providing server 1140 acquires the access history of these users. The primary collection location of the log is an information service providing server 1140 (web site) that acquires this access history.

  For log collection, for example, the access history of each information page and product page of a website that provides information and products is used. In general, there are a plurality of websites that hold access histories. In this case, accesses representing the same information and the same product must be recorded with the same name. In addition, the user identifier needs to be able to confirm that it is the same user even if it crosses each site.

  In this embodiment, it is assumed that log information of the information service providing server 1140 is collected in the log management server 1130 through the network 1150 and can be managed collectively. User identifiers are managed uniformly.

  In some cases, the log management server 1130 and the information service providing server 1140 may be the same. For example, in a Web directory service that provides information on each Web site at a certain site, which site the user has selected and moved is collected as an access history. If the log management server is also running in the site, the access history can be passed without going through the network. If the access history can be acquired from the information service providing server and the relay server of the user terminal, the log can also be used.

  On the other hand, sites that provide large-scale services often have a plurality of service providing servers. In this case, the access history at each server is sent to a log management server at one location via the network for management. The same applies when the service history is held by a network provider such as an ISP.

  The user identifier is usually assigned by the service providing server 1140. When assigning a uniform identifier among a plurality of sites, there are cases where the sites are unified, or a unique identifier is assigned by inquiring to the log management server 1130. It is done.

  The log management server 1130 sends the access history of each user to the advertisement providing server 1120 through the network 1150 automatically at regular intervals or in response to a request from the advertisement providing server 1120. In the advertisement providing server 1120, the user history management unit 1122 manages the user access history sent from the log management server 1130. As a result, the user history management unit 1122 manages the latest access history. FIG. 20 shows an example of a user access history managed by the user history management unit 1122.

  Note that the advertisement providing server 1120 itself may also serve as the log management server 1130. Further, the advertisement providing server 1120 itself may be one of information service providing servers.

  Next, processing for providing an advertisement using the user access history in the advertisement providing server 1120 will be described.

  FIG. 12 shows the overall flow of the advertisement providing process. When an advertiser makes an advertisement posting request, a latent target information list is generated and an advertisement agent is set, and advertisements, additional information, etc. are provided to the user. In addition, the number is counted and reported to the advertiser, and payment is received.

  In the following, the flow of processing will be described in detail with reference to FIG. The advertisement provided to the user is referred to as advertisement information, and information that can be referred to by the user is referred to as reference information.

  (1) The advertiser uses the advertiser terminal 1100 to send the advertisement information to be submitted to the advertisement providing server 1120. At that time, the contents of the advertisement, the submission period, the location where the additional information is saved, etc. are also notified. In the advertisement providing server 1120, the advertisement information sent from the advertiser terminal 1100 is managed by the advertisement information management unit 1121. FIG. 21 shows an example of advertisement information managed by the advertisement information management unit 1121.

  Further, the advertiser selects one or more target information (website information page, product page, etc.) related to the advertisement information to be posted and sends it to the advertisement providing server 1120 when requesting the advertisement posting. To do. This target information may be selected and notified by an advertiser's agent, agency, or the like. FIG. 22 shows an example of the target information.

  (2) The latent target information generating unit 1123 of the advertisement providing server 1120 extracts the latent target information that is G3 when the target information is G1 in FIG. FIG. 14 is a conceptual diagram of the extraction of latent target information, and FIG. 15 is a process flow diagram of the extraction of latent target information.

  First, a group G2 of users (related users) referring to the target information G1 is extracted from the collected user access history (step 1501). Next, if the number of extracted users (number of related users) is large, the number is reduced to a certain number (step 1502). This is performed, for example, by checking the number of times the target information has been referenced for each extracted user (step 1502-1) and narrowing down to the fixed number in order from the largest number of times of reference (step 1502-2).

  Next, from the collected user access history, a group of information other than the target information G1 is collected from the information referred to by each extracted user (related user) (step 1503). This is the latent target information G3 for the target information G1. Again, if there are many collected information groups (latent target information groups), the number is reduced to a certain number (step 1504). This method is the same as in the case of reducing the number of related users (steps 1504-1 and 1504-2). FIG. 23 shows an example of latent target information.

  (3) The user history management unit 1123 of the advertisement providing server 1120 sets a virtual user (referred to as an advertisement agent) using the additional information, the target information, and the latent target information as reference information, and the user access history. Add to That is, it is as if the advertising agent referred to additional information, target information, or latent target information. Here, a virtual fixed user identifier is assigned to the advertising agent. At that time, an advertisement attribute is attached to the advertisement agent, and an advertisement attribute is also attached to the additional information.

  FIG. 24 shows an example of the virtual access history of the advertising agent obtained from FIGS. FIG. 25 shows an example of attribute settings for the user identifier and the reference information identifier. In some cases, the latent target information may be omitted, and the advertising agent may use only the additional information and the target information as reference information, or both the target information and the latent target information and the additional information may be used as reference information. Good.

  (4) The advertisement information selection unit 1124 of the advertisement providing server 1120 selects advertisement information most suitable for the user based on the algorithm shown in FIG. However, the users appearing in G2 are only those who have the advertisement attribute, that is, the advertising agent, and the introduction information appearing in G3 is also the information having the advertisement attribute, that is, only the additional information. The advertisement content is obtained from the additional information, and the advertisement content and the additional information are linked to provide the advertisement content to the user. FIG. 16 is a conceptual diagram of the advertisement posting processing algorithm, and FIG. 17 is a processing flowchart.

  First, information (reference information) G1 referred to by the user of the advertisement providing partner is extracted from all user access histories (including virtual advertisement agent access histories) (step 1701). Next, a group G2 of users (advertising agents) that refer to the reference information and have advertising attributes is extracted (step 1702). Specifically, a group of users (related users) other than the advertisement providing user who refers to the reference information G1 is extracted from the entire access history, and the user group having the advertisement attribute is extracted from the user group. A group of (advertising agents) is extracted. Next, when the number of extracted users (advertising agents) G2 is large, it is reduced to a certain number (step 1703).

This is performed, for example, by checking the number of times the reference information is referred to for each extracted user (step 1703-1), and narrowing down to a certain number in order from the person with the highest number of references (step 1703-2).

  Next, a group G3 of information (additional information) with advertisement attributes referred to by each user (advertisement agent) with the extracted advertisement attributes is extracted (step 1704). Specifically, an information group referred to by the advertising agent G2 is extracted from the all user access history, and information (additional information) with an advertisement attribute is extracted from the information group. Again, if there are many extracted additional information groups, the number is reduced to a certain number (for example, the number necessary for display) (steps 1705, 1705-1, and 1705-2).

  Next, the advertisement content associated with the extracted additional information is obtained, the additional information is linked to the advertisement content, and the advertisement content is sent to the user terminal 1110 of the user of the advertisement providing partner (step 1706).

  In this way, optimal advertisement information is provided to the user according to the reference information (access history) of each user. The advertisement information can be provided to the user, for example, when requested by the user, or from the advertisement providing server 1120 at regular intervals.

  (5) The user refers to the advertisement content using the browser of the user terminal 1110, and if interested, requests additional information from the advertisement providing server 1120 and acquires the additional information. FIG. 18 shows a specific image thereof.

  (6) The advertisement provision number management unit 1125 of the advertisement provision server 1120 records the number of times the user has referred to the advertisement and the number of times the additional information has been acquired. In addition, the advertisement fee calculation unit 1126 calculates the advertisement fee based on the advertisement reference count and the additional information acquisition count.

  The advertisement providing server 1120 makes a predetermined report to the advertiser through the advertiser terminal 1100 based on the advertisement reference record, the additional information acquisition record, and the charge calculation result as needed, or after the advertisement posting period ends, and pays the advertisement price. To charge.

FIG. 19 shows various methods for paying the advertisement. (1) in FIG. 19 is a method of determining the advertisement price at the stage when the advertisement posting is determined. Such can also be used in conventional while the payment method, the present invention is suitable for the model to determine the advertisement price, especially in light of the provisions the number of ad reference times the number and additional information acquisition times. (2) to (4) in FIG. 19 are examples thereof.

(2) of FIG. 19, advertising the number of times the advertising reference times the number or additional information acquired at the time of posting or pay in advance by performing the minimum number guaranteed for both, if you did not reach it, advertisement posting period to reach it It is a method of extending.

In addition, (3) of FIG. 19, advertising the number of times the advertising reference times the number or additional information acquired at the time of posting or pay in advance by performing the minimum number guaranteed for both, if you did not reach it, on the basis of the provisions This is a method of refunding part or all of the charge equivalent to the difference.

Further, (4) in FIG. 19, the aggregated advertising reference times speed and based on the number of additional information acquisition times, a method for performing a payment after posting period.

  Specific examples are shown below. Here, a banner advertisement on the WWW will be described as an example on the premise of a web service providing product information on the Internet. It is assumed that the user is referring to various Web pages using this Web service.

  The advertiser or his agent prepares a banner advertisement for a new camera-equipped personal computer, and requests the advertisement to be posted. At that time, the URL of advertisement information (advertisement image, catch phrase) and URL of additional information further referred to by users who are interested in the advertisement are set. Usually, it is set as link information for this advertisement content. FIG. 26 shows a specific example of advertisement information.

  Further, the advertiser or agent selects several web pages (target information) related to the advertisement from the web service pages. Here, it is assumed that the Web page in FIG. 27 is selected.

  The advertisement providing server 1120 first extracts a page (latent target information) related to the selected Web page. That is, the web page group in FIG. 27 is set to G1, and after extracting related users of G2, latent target information is extracted as G3. The history used when extracting G1 to G2 and G2 to G3 here is a log accessed in this Web service. A specific example of the latent target information is shown in FIG.

  When the target information is recorded in the access history of the advertisement agent as information directly related to the advertisement recorded as the additional information, the advertisement can be sent to the client who has accessed the target information. Similarly, when latent target information is recorded in the access history of an advertising agent, an advertisement can be sent to a client who has not accessed target information but is accessing latent related information related to the target information. That is, the advertisement can be distributed to clients who are potentially interested in the advertisement.

  Next, a user ID as an advertising agent is assigned, and a record (access history) that the user ID refers to the additional information URL, the target information URL, and the latent target information URL is left. Here, the access history of the advertising agent is as shown in FIG. At that time, as shown in FIG. 30, the advertisement attribute is set in the user ID of the advertisement agent, and the advertisement attribute is set in the additional information URL.

  When a general user tries to refer to an advertisement when using a Web service, the advertisement providing server 1120 operates to introduce the advertisement. At that time, the user who enters G2 is only the user of the advertisement attribute, that is, the advertisement agent, and the user who enters G3 is also the one of the advertisement attribute, that is, only the additional information. The advertisement content is obtained from the extracted additional information. Therefore, only the advertisement is introduced.

  The user checks the displayed advertisement and acquires additional information if necessary. Specifically, the advertisement image is clicked as shown in FIG. The advertisement providing server 1120 also stores the access record.

  Regarding payment of charges, several cases are prepared as shown in FIG. Generally, a fee is paid at the time of an advertisement posting request, but three other options are prepared here.

  If the guaranteed number of advertisement references and the number of additional information acquisitions are not exceeded, the first option for posting until reaching the target is the charge table as shown in FIG. 31 and the guaranteed number of advertisement references and additional information acquisition as shown in FIG. Since the number did not exceed, the posting period was delayed as shown in FIG. 33 and posted until reaching the guaranteed number of advertisement references and the number of additional information acquisition.

In this example, the posting period was 4 / 1-30, but it was delayed until 5/15.

  If the guaranteed number of advertisement references and additional information acquisition number are not exceeded, the second option for refunding the amount is the fee table as shown in FIG. 31 and the guaranteed advertisement reference number and additional information acquisition as shown in FIG. I did not exceed the number and refunded it. In this example, 71000 yen (rounded down to the nearest 1000 yen) was refunded.

  In the third option for determining the fee based on the actual number of advertisement references and the number of additional information acquisitions, the unit price for the number of advertisement postings is 3 yen and the unit price for the number of additional information acquisitions is 100 yen in the fee table as shown in FIG. As a result of posting up to 1-30, as shown in FIG. 36, it was calculated based on the actual number of advertisement references and the number of additional information obtained, and 328,000 yen (rounded to 1000 yen or less) was charged.

  According to this embodiment, a user who is likely to be interested and a user who is likely to be interested in advertising in interactive media is discovered, and an appropriate information is obtained based on the latest access log of the user. Advertisement information can be selected and provided. Thereby, in the advertising fee system based on the number of advertisement references and the number of additional information acquisition, even if the number of advertisement references is the same, the number of additional information acquisition increases, and there is an advantage of obtaining more income.

  Further, according to this embodiment, if a plurality of pieces of target information are set, it is possible to target a user who refers to a plurality of specific contents rather than a user who refers to only one specific content. In addition to target information, the target can be expanded by determining the target users based on latent target information.

  Next, a third embodiment will be described in detail with reference to FIGS.

  37 and 38 are block diagrams showing the configuration of the advertisement distribution system according to this embodiment.

  The client side shown in FIGS. 37 and 38 has an input / output device 61 for browsing item information, purchasing items, and displaying advertisements, and a customer ID storage unit 62 for holding a customer ID. A customer-accessible terminal 60 such as a personal computer or other Web-accessible device, a digital interactive TV set, a telephone, or a mobile phone.

  The input / output device 61 includes a display, a printer, a television screen, a liquid crystal screen, and the like as an output unit, and includes a mouse, various pointing devices, a keyboard, an infrared remote controller, a mobile phone push key, and the like as an input unit.

  The customer ID storage unit 62 stores the customer ID allocated from the online shopping server 10.

  On the other hand, in response to a request from the customer terminal 60, the server side extracts potential customers of items to be distributed from the display of item information, the online shopping server 10 that receives purchase orders for items, and the access history of the online shopping server 10 Potential customer extraction server 20, advertisement transmission server 30 that displays or distributes advertisements to customer terminal 60, and billing information management server 40 that manages billing information for advertisement distribution.

  The customer terminal 60, online shopping server 10, potential customer extraction server 20, advertisement transmission server 30, and billing information management server 40 are connected by a network N such as the Internet.

  Online shopping server 10 is available at Amazon.com (http://www.amazon.com/), Outpost.com (http://www.outpost.com/), Rakuten Ichiba (http: //www.rakuten .com.jp /) and other EC sites that can be used to view item information and actually purchase items, and receive item information browsing requests and purchase requests from customer terminals 60. An online shopping processing unit 12 that performs information display and purchase order order processing, an access history list 50 that records access (view / purchase) history of items such as browsing / purchasing of the items, and a customer ID to the customer terminal 60 Issued and configured with an ID information management unit 13 that stores customer ID information in the customer data storage unit 14, an item data storage unit 15 that stores item information and item ID, and a control unit 11 that controls each unit Is done.

  The item is information or a product that can be browsed, held in a basket, etc., purchased through the customer terminal 60, and can be uniquely identified by an item ID such as a Web URL or a product code. .

  The customer ID is a globally unique identifier, and the ID information management unit 13 newly generates or confirms the existing ID.

  The potential customer extraction server 20 includes an access history collection unit 22 that collects an access history list 50 from one or a plurality of online shopping servers 10, and an access history list storage unit that holds the collected access history list 50 for a certain period of time. 23, a potential customer extraction processing unit 24 that extracts a potential customer in accordance with a potential customer extraction request received from the input unit 25 based on the history of the access history list storage unit 23, and the extracted list of potential customers Output unit 26 and control unit 21 for controlling them.

  The advertisement transmission server 30 outputs the designated advertisement target item ID to the potential customer extraction server 20, the input unit 33 that receives the potential customer list, and the customer terminal of the customer ID described in the potential customer list. It consists of an advertisement distribution processing unit 32 that transmits advertisements, a customer data storage unit 35, an item data storage unit 36 similar to the customer data storage unit 14 of the online shopping server 10, and a control unit 31 that controls these. The

  The billing information management server 40 is charged as a trigger for sending advertisements, responding to advertisements (clicking on banners, etc.), browsing item information, purchasing items, etc. Is a management server for charging a fee, and includes a charging condition management unit 42, an advertisement distribution frequency management unit 43, an advertisement response management unit 44, an input unit 45, an output unit 46, and a control unit 41.

  The billing condition management unit 42 manages the advertisement billing type, the number of transactions (the number of advertisement distributions, the number of advertisement responses), and the contract fee for each item ID, and the number of transactions (the number of advertisement distributions, The number of advertisement responses) and the contract fee can be read out.

  The following types of advertisement billing are possible.

  -Pay-per-use billing for each number of advertisement delivery: Number of advertisement delivery of contract target item x charge per delivery.

  -Flat-rate billing for the predetermined number of advertisement distributions: Charges on condition that the number of advertisement distributions for the contract target item reaches a predetermined number.

  -Pay-per-use billing for each number of advertisement responses: Number of responses to advertisement distribution of contracted items x Charge per response.

  -Flat-rate billing for the number of predetermined advertisement responses: A charge on condition that the number of responses to the advertisement distribution of the contract target item reaches a predetermined number.

  ・ Billing with the above combination: Set a fixed amount of flat fee on condition that a certain number of advertisements are delivered Refund the amount.

  FIG. 39 is a diagram showing an access history list 50 generated by the online shopping server 10 of the advertisement distribution system according to the third embodiment.

  The access history list 50 is held in a format including at least “access customer ID 51, accessed item ID 52”, and has additional information such as access date and time 53, for example, means for using the access history of the Web server, There is a means by a unique history holding means.

  FIG. 41 is a diagram showing an advertisement target item list 80 input to the potential customer extraction server 20 of the advertisement distribution system according to the third embodiment.

  The advertisement target item list 80 is described in a format including at least “advertisement target item ID 81, priority 82”. For example, a description method using a comma-delimited CSV file format or a description method using a unique description format is available. .

  In the advertisement target item list 80, an item ID of an item related to the item may be described in addition to an item ID 81 of an item to be advertised.

  FIG. 42 is a diagram showing a potential customer list 90 output from the potential customer extraction server 20 of the advertisement distribution system according to the third embodiment.

  The potential customer list 90 is described in a format including at least “priority 91 and customer ID 92”. For example, there are means for describing in a comma-delimited CSV file format and a description method using a unique description format.

  Next, detailed operation of each part will be described with reference to FIGS.

<Grant customer ID>
FIG. 43 is a flowchart showing a customer ID assigning operation in the online shopping server 10.

(1) A customer uses the customer terminal 60 to access the online shopping server 10 via the network 11 such as the Internet. (Step 2100)
(2) Upon confirming the connection to the online shopping server 10, the customer terminal 60 sends the customer ID stored in the customer ID storage unit 62 together with the “customer terminal authentication request”. At this time, when the customer ID is not held, a NULL code indicating nothing is transmitted. (Step 2110)
(3-1) The online shopping server 10 that has received the “customer terminal authentication request” and the customer ID, if the customer ID is a NULL code (step 2120), is newly globally unique from the ID information management unit 13. A customer ID, which is a unique identifier, is generated (step 2140), the customer ID is stored in the customer data storage unit 14 (step 2150), and the customer ID is sent to the customer terminal 60 together with a “customer terminal authentication response” (step 2150). 2160).

  (3-2) The online shopping server 10 that has received the “customer terminal authentication request” and the customer ID makes an inquiry to the ID information management unit 13 when the customer ID is not a NULL code (step 2120), and the customer data storage unit 14 If the customer ID does not exist (step 2130), the customer ID received from the customer terminal 60 is discarded, and the ID information management unit 13 newly adds a globally unique ID. A customer ID as an identifier is generated (step 2140), the customer ID is stored in the customer data storage unit 14 (step 2150), and the customer ID is sent to the customer terminal 60 together with a “customer terminal authentication response” (step 2160). .

  (3-3) The online shopping server 10 that has received the “customer terminal authentication request” and the customer ID makes an inquiry to the ID information management unit 13 when the customer ID is not a NULL code (step 2120), and the customer data storage unit 14 If it is an existing customer ID (step 2130), the customer ID is sent to the customer terminal 60 together with a “customer terminal authentication response” (step 2160).

  (4) Upon receiving the “customer terminal authentication response” and the customer ID, the customer terminal 60 writes the received customer ID into the customer ID storage unit 62 (step 2170).

  In this series of processing, the ID information management unit 13 stores and manages the password together with the customer ID, thereby performing password check for the processing of the customer authentication request of (2) to (3) and increasing the security of the customer ID. Is also possible.

<Generation of access history>
FIG. 44 is a flowchart showing an access history generation operation in the online shopping server 10.

  (1) The customer uses the customer terminal 60 to browse the item information on the online shopping server 10 or transmit an item purchase request via the network N such as the Internet (step 2200).

  (2) The access history is added to the access history list 50 so as to include at least the customer ID of the customer terminal that has accessed and the item ID of the item to be browsed / purchased (step 2210).

  The information recorded on the access history is not limited to the above, and information such as browsing / purchasing distinction can be recorded and used as a prioritization parameter in the potential customer extraction operation described later. is there.

<Collecting access history>
FIG. 45 is a flowchart showing an access history collection operation in the potential customer extraction server 20.

  (1) In the potential customer extraction server 20, an access history collection request is transmitted from the control unit 21 to the access history collection unit 22 at a predesignated time, and an access history collection operation is started (step 2300).

  (2) In the potential customer extraction server 20, the access history collection unit 22 collects the access history list 50 on the online shopping server 10 using a protocol such as FTP GET via the network N (step 2310).

  (3) The access history collection unit 22 that has collected the access history list 50 from the online shopping server 10 holds it in the access history list storage unit 23. At this time, the access history browsed / purchased by the customer using the customer ID as a key and the access history of browsing / purchased the item by using the item ID as a key are stored (step 2320).

  In the collecting operation (2), the number of online shopping servers 10 to be collected is not limited to one, and may be collected from a plurality of servers 10. Further, the method is not limited to the method of collecting the access history from the access history collecting unit 22, but may be a method of transmitting from the online shopping server 10 to the access history collecting unit 22 periodically or at the time of access.

<Potential customer extraction>
FIG. 46 is a flowchart showing an operation of extracting a potential customer in the potential customer extraction server 20.

  (1) In the potential customer extraction server 20, the input unit 25 receives the input of the advertisement target item list together with the “potential customer extraction request” (step 2400).

  The information of the advertisement target item list is not limited to only the item ID of the item that is actually the target of the advertisement, and may include the item ID of the item related to the advertisement target item specified by the advertiser.

  (2) The potential customer extraction request is transmitted from the input unit 25 that has received the potential customer extraction request to the control unit 21, and the potential customer extraction operation is started (step 2410).

  (3) A latent customer extraction start request is transmitted from the control unit 21 that has received the latent customer extraction request to the latent customer extraction processing unit 24 (step 2420).

  (4) Upon receiving the latent customer extraction start request, the potential customer extraction processing unit 24 reads the advertisement target item list from the input unit 25, and sets all item IDs in the advertisement target item list as the target item group 71 (hereinafter referred to as G1). (Refer to FIG. 40) (step 2430).

  (5) In the potential customer extraction processing unit 24, the access history list storage unit 23 obtains an access history in which the customer has viewed / purchased the item in the past using the item ID of the G1 as a key. The customer ID is set in the customer (direct customer) group 72 (hereinafter referred to as G2: see FIG. 40) who has accessed the item of G1 (step 2440).

  (6) In the potential customer extraction processing unit 24, in the same manner as 5), from the access history list storage unit 23, an access history in which the customer has viewed / purchased items in the past is stored using the customer ID in the G2 as a key. It is acquired and set in the item group 73 (hereinafter referred to as G3: see FIG. 40) closely related to the G1 item that is not G1 among all the item IDs in the access history (step 2450).

  (7) In the potential customer extraction processing unit 24, in the same manner as in 5), the access history that the customer has viewed / purchased the item in the past is acquired from the access history list storage unit 23 using the item ID in G3 as a key. Then, out of all customer IDs in the access history, a customer ID that is not in G2 is assigned to a customer (potential customer) group 74 (hereinafter referred to as G4: see FIG. 4) that is likely to be interested in the item of G1. Set (step 2460).

  (8) The potential customer extraction processing unit 24 transmits the G4 customer ID as a potential customer list to the output unit 26 (step 2470).

  (9) The output unit 26 that has received the potential customer list outputs the potential customer list to the external interface (step 2480).

  By the above operation, it is possible to extract a customer group having an access tendency similar to that of the customer group who has accessed the target item to be advertised. Since it can be said that the customer group having the access tendency similar to the direct customer has the similar interest tendency, that is, the customer group (potential customer) having a high possibility of purchasing the advertisement target item is extracted. become.

  In the processes (4) to (7), for example, the number of accessed items (overlap amount) in the group, the access date / time, the number of past access items of the customer, the number of past accessed customers of the item, etc. By sorting the order as a key, the priority is determined, and by narrowing down to the top n items of the sorted order, it is possible to improve the accuracy of the similarity of the access tendency.

<Advertising delivery operation>
FIG. 47 is a flowchart showing an operation of distributing advertisements to potential customers in the advertisement transmission server 30.

  (1) In the advertisement transmission server 30, the input unit 33 receives an input of one or a plurality of advertisement target item IDs together with the “advertisement distribution request” (step 2500).

  (2) The advertisement transmission server 30 transmits the advertisement distribution request from the input unit 33 to the control unit 31, and the advertisement distribution processing operation starts (step 2510).

  (3) The control unit 31 that has received the advertisement distribution request transmits an “advertisement target item ID output request” to the output unit 34 (step 2520).

  (4) Upon receiving the advertisement target item ID output request, the output unit 34 reads the advertisement target item ID from the input unit 33 (step 2530).

  (5) The advertisement target item ID is transmitted together with the “latent customer extraction request” from the output unit to the latent customer extraction server 20 (step 2540).

  (6) The potential customer extraction server 30 executes a potential customer extraction operation for the item to be advertised (step 2550 / steps 2400 to 2480).

  (7) The potential customer list output from the potential customer extraction server 20 is received by the input unit 33 of the advertisement transmission server 30 (step 2560).

  (8) The latent customer list is transmitted from the input unit 33 of the advertisement transmission server 30 that has received the latent customer list to the advertisement distribution processing unit 32 (step 2570).

  (9) An advertisement is transmitted from the advertisement distribution processing unit 32 that has received the potential customer list to the customer terminal 60 corresponding to the customer ID described in the potential customer list using means such as an e-mail or a push distribution device. (Step 2580).

  The advertisement distribution method is not limited to a method of transmitting directly using an e-mail or a push distribution device, and may be a method of displaying a banner advertisement when the customer terminal 60 accesses the online shopping server 10. .

According to this embodiment, since the advertisement distribution destination can be narrowed down to potential customers based only on the past access history of the customer,
-No need to register attribute information in advance from customers (easy to collect customers because there is no need to input attributes)
・ Registration of attribute information for items is not required (no maintenance work is required)
・ Advertisement can be delivered to potential customers of the target item, and the customer is more likely to be interested in the target item. In other words, the customer's labor and management are reduced, and the target item can be viewed. / The probability of purchase can be increased.

  The present invention is not limited to the above. For example, the advertisement distribution can be displayed on the top page of the online shopping server 10 in the form of a recommended item.

  In addition, it is effective not only for Internet applications but also for determining advertisement distribution destinations based only on access history in media and networks where it is difficult to input attributes such as digital interactive TV and information appliances.

  The online shopping server 10, the potential customer extraction server 20, the advertisement transmission server 30, and the billing information management server 40 may be independent devices, or any combination or all may be implemented by a single server.

  As described above, in an online shopping server such as a Web EC site, a history of a customer viewing item information or purchasing an item (access history) is stored. The stored information is, for example, “customer ID, item ID, access date”.

  When the above access history is collected and the advertisement of item A is distributed, as shown in FIG. 4, “customer group G2 who browsed / purchased item group G1 consisting of item A and related items browsed / purchased in common. The basic logic is to distribute the advertisement to the “customer (potential customer) group G4 who is likely to view / purchase G1w” who sees the item group G3 in the same manner.

  In this way, according to this embodiment, an advertisement for a customer who is likely to view / purchase the target item to which the advertisement is to be distributed based only on the access history (no attribute pre-registration is required). Distribution is possible.

  It will be appreciated by those skilled in the computer art that the above embodiments may be conveniently implemented using a conventional general purpose digital computer programmed according to the teachings herein and disclosed by a programmer. Appropriate software code can be easily created based on the teachings of.

  In particular, various server devices such as the information delivery server, the advertisement providing server, the online shopping server, the potential customer extraction server, and the advertisement transfer server in each of the above embodiments can be provided in a convenient form in software form.

  Such a software package may be a computer program product that uses a recording medium containing recorded computer code that is used to program the computer to perform the disclosed functions and processes of the present invention. The recording medium may be a conventional floppy disk, optical disk, CD-ROM, magneto-optical disk, ROM, RAM, EPROM, EEPROM, magnetic or optical card or other medium suitable for storing electronic instructions. However, it is not limited to these.

It is a system configuration figure of a first embodiment. It is a processing flow figure of access history acquisition in a first embodiment. It is a registration processing flow figure of the correlation item delivery destination client in 1st embodiment. It is a registration processing flow figure of correlation item delivery in a first embodiment. It is a whole processing flow figure of correlation item detection in a first embodiment. It is a conceptual diagram of the correlation item detection in 1st embodiment. It is a detailed process flow figure of correlation item detection in a first embodiment. It is a processing flow figure of delivery and charge of a correlation item in a first embodiment. It is a totalization processing flow figure of the billing information in a first embodiment. It is a figure which shows the specific example of the correlation item delivery in 1st embodiment. It is a system configuration figure of a second embodiment. It is a figure which shows the flow of the whole process of the advertisement introduction in 2nd embodiment. It is a conceptual diagram of the social filtering information introduction algorithm in 2nd embodiment. It is a conceptual diagram of the latent target information extraction algorithm in 2nd embodiment. It is a flowchart of the process of the latent target information extraction in 2nd embodiment. It is a conceptual diagram of the algorithm of the advertisement introduction in 2nd embodiment. It is a processing flow figure of advertisement introduction in a second embodiment. It is a figure which shows the specific image of the advertisement introduction in 2nd embodiment. It is a figure which shows the type of the payment method in 2nd embodiment. It is an example of the access history in 2nd embodiment. It is an example of the advertising information in the second embodiment. It is an example of the object information in 2nd embodiment. It is an example of the latent object information in 2nd embodiment. It is an example of the access log | history creation of the advertising agent in 2nd embodiment. It is an example of the attribute setting of a user identifier and information in 2nd embodiment. It is a specific example of the advertisement information in the second embodiment. It is a specific example of object information in a second embodiment. It is a specific example of the latent target information in the second embodiment. It is a specific example of the access history creation of the advertising agent in the second embodiment. It is a specific example of a user identifier and information attribute setting in the second embodiment. It is a specific example of the guaranteed number of times at the time of contract in the second embodiment. It is a specific example of the report after the advertisement posting period in the second embodiment. It is a specific example of the report after the additional posting in 2nd embodiment. It is another specific example of the report after the advertisement posting period in the second embodiment. It is a specific example of the unit price at the time of contract in the second embodiment. It is another specific example of the report after the advertisement posting period in the second embodiment. It is a block diagram which shows the structure of the advertisement delivery apparatus which concerns on 3rd embodiment. It is a block diagram which shows the structure of the advertisement delivery apparatus which concerns on 3rd embodiment. It is explanatory drawing which shows the access history list produced | generated with the online shopping server of the advertisement delivery apparatus which concerns on 3rd embodiment. It is a conceptual diagram which shows an example of the potential customer extraction which concerns on 3rd embodiment. It is explanatory drawing which shows the advertisement object item list input into the potential customer extraction server of the advertisement delivery apparatus which concerns on 3rd embodiment. It is explanatory drawing which shows the potential customer list output from the potential customer extraction server of the advertisement delivery apparatus which concerns on 3rd embodiment. It is a flowchart which shows customer ID provision operation | movement in the online shopping server which concerns on 3rd embodiment. It is a flowchart which shows the access log generation operation | movement in the online shopping server which concerns on 3rd embodiment. It is a flowchart which shows the access history collection operation | movement in the potential customer extraction server which concerns on 3rd embodiment. It is a flowchart which shows operation | movement of the potential customer extraction in the potential customer extraction server which concerns on 3rd embodiment. It is a flowchart which shows the operation | movement which distributes the advertisement to a potential customer in the advertisement transmission server which concerns on 3rd embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Online shopping server 11 Control part 12 Online shopping process part 13 ID information management part 14 Customer data storage part 15 Item data storage part 20 Potential customer extraction server 21 Control part 22 Access history collection part 23 Access history list storage part 24 Potential customer extraction Processing unit 25 Input unit 26 Output unit 30 Advertisement transmission server 31 Control unit 32 Advertisement distribution processing unit 33 Input unit 34 Output unit 35 Customer data storage unit 36 Item data storage unit 40 Billing information management server 41 Control unit 42 Billing condition management unit 43 Advertisement distribution frequency management unit 44 Advertisement response management unit 45 Input unit 46 Output unit 50 Access history list 60 Customer terminal 61 Input / output device 62 Customer ID storage unit 100 Client 110 Information delivery server 111 Web page 112 Server Engine 113 Inter-item correlation detection unit 114 Browsing / purchase history database 115 Client management unit 116 Base item holding unit 117 Correlated item holding unit 118 Charge management unit 120 Information browsing server 130 Network 1100 Advertiser terminal 1110 User terminal 1120 Advertising providing server 1121 Advertisement information management unit 1122 User history management unit 1123 Potential target information generation unit 1124 Advertisement information selection unit 1125 Advertisement provision frequency management unit 1126 Advertisement fee calculation unit 1130 Log management server 1140 Information server provision server 1150 Network

Claims (9)

The advertisement providing device
Access history of user access to information provided by the information provider and additional information accessible to the client that is advertisement information to be provided to the client who is the user accessing the target information specified in advance. a step of the advertisement information associated with and stored in the data base over scan manages and,
A history that a virtual user has accessed both the target information and the additional information is stored in the database as the access history by associating the unique identifier for identifying the virtual user with the target information and the additional information. a step of causing further store,
Based on the access history, the advertisement information selection means detects a unique identifier associated with the target information, with a virtual user having a history of accessing the target information other than the client as a related user, As the correlation information accessed by the related user based on the access history, the additional information associated with the unique identifier is extracted, and the advertisement managed in association with the extracted additional information Delivering advertisements in response to requests from clients by delivering information to clients;
An advertisement providing method characterized by comprising: Latent target information generating means detects other related users who have accessed the target information other than the client and the virtual user based on the access history;
Latent target information generating means extracts latent target information accessed by the other related users based on the access history; and
A step of another virtual user is a history that has access both the additional information and the target information, as said access history, Ru is stored in the database,
Based on the access history, the advertisement information selection means, for another client accessing the latent target information, as another related user who accessed the potential target information other than the other client, Advertising information managed in association with the extracted additional information, by detecting the user and extracting the additional information as correlation information accessed by the other related user based on the access history Delivering to another client;
The advertisement providing method according to claim 1, further comprising:   The advertisement providing method according to claim 1, wherein the providing step provides the advertisement information in a form linked to the additional information.   The advertisement providing method according to claim 1, wherein the managing step manages the advertisement information designated by an advertiser, the additional information, and the target information. A step in which the advertisement provision number management means totals the number of advertisement information delivery and the number of additional information accesses by the actual advertisement provision operation and stores them in the database;
Advertising rate calculation means, the steps of charging advertisers advertising fees based on at least one of said advertisement information delivery count and the additional information access count,
The advertisement providing method according to claim 1, further comprising: Advertising fee calculation means, so that the payment of prior advertising fee to the advertiser the actual ad serving behavior, have been guaranteed is stored in the data base over the scan lowest advertisement information delivery count and the minimum additional information access count Charging the advertiser with an advertising fee for at least one of them ,
A step in which the advertisement provision number management means totals the number of advertisement information delivery and the number of additional information accesses by the actual advertisement provision operation and stores them in the database;
The advertisement provision frequency management means until the advertisement information delivery count reaches the minimum advertisement information delivery count, or until the additional information access count reaches the minimum additional information access count, or the advertisement information delivery count a step but which reached the lowest advertisement information delivery times, and to the additional information access count reaches the minimum additional information access count, extending the advertisement posting period for the actual advertisement providing operation,
The advertisement providing method according to claim 1, further comprising: Advertising rate calculation means, so that the payment of prior advertisement fee to an advertiser actual advertisement providing operation, at least one of minimum advertisement information delivery count and minimum additional information access count that is guaranteed is stored in the database Charging advertisers for one-side advertising fees,
A step in which the advertisement provision number managing means totals the number of advertisement information delivery times and the number of additional information accesses by the actual advertisement provision operation in a predetermined advertisement posting period and stores them in the database ;
A part of the advertising fee corresponding to at least one of the difference between the advertisement information delivery frequency and the minimum advertisement information delivery frequency and the difference between the additional information access frequency and the minimum additional information access frequency. At least a refund step,
The advertisement providing method according to claim 1, further comprising: Access history of user access to information provided by the information provider and additional information accessible to the client that is advertisement information to be provided to the client who is the user accessing the target information specified in advance. and means for managing by stores and the advertising information in a database associated with,
A history that a virtual user has accessed both the target information and the additional information is stored in the database as the access history by associating the unique identifier for identifying the virtual user with the target information and the additional information. It means for causing further store,
Based on the access history, a virtual user who has a history of accessing the target information other than the client is regarded as a related user , and a unique identifier associated with the target information is detected. Based on the access history As the correlation information accessed by the related user, the additional information associated with the unique identifier is extracted, and the advertisement information managed in association with the extracted additional information is delivered to the client. Means to provide advertisements in response to requests from clients,
An advertisement providing apparatus comprising: Access history of user access to information provided by the information provider and additional information accessible to the client that is advertisement information to be provided to the client who is the user accessing the target information specified in advance. a step of managing stores the advertisement information in the database associated with,
A history that a virtual user has accessed both the target information and the additional information is stored in the database as the access history by associating the unique identifier for identifying the virtual user with the target information and the additional information. A further storing step ;
Based on the access history, a virtual user who has a history of accessing the target information other than the client is regarded as a related user , and a unique identifier associated with the target information is detected. Based on the access history As the correlation information accessed by the related user, the additional information associated with the unique identifier is extracted, and the advertisement information managed in association with the extracted additional information is delivered to the client. by the step of providing an advertisement in response to a request from a client,
A computer program for providing advertisements, characterized by causing a computer to execute .

Images