JP2004094383A - Recommendation device and advertisement distribution method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a recommendation device for enhancing reliability of taste information on a user being an object for recommending the contents. <P>SOLUTION: A history of the contents transmitted to a user terminal and a category set for classifying the contents are associatively stored in a user behavior history database. The taste information on respective users is generated by making a classification-analysis on the basis of a distribution history of these contents, and is stored in a user database. When distributing an advertisement of the contents, the advertisement is distributed according to a hobby and a taste of the user on the basis of the taste information stored in this user database. The distribution history of the contents stored in the user behavior history database is correlatively analyzed or discriminatingly analyzed, and is stored in an analytical result database, and the advertisement of the contents is distributed to the user terminal on the basis of this analysis. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a recommendation device that distributes an advertisement to a user terminal.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, various services using a recommendation device have been provided. For example, the user can answer a questionnaire about hobbies and preferences, register the result of the questionnaire as preference information in a server, and transmit an advertisement of recommended content according to the preference information of the user to each user terminal. . Then, the user refers to these advertisements, makes a transmission request for necessary content, and purchases the content. In this questionnaire, for example, the user selects a content category (travel, music, sports, etc.) according to his / her own hobbies and preferences, and further selects his / her preference from among five levels. Registered as information.
[0003]
[Problems to be solved by the invention]
However, according to the above-described related art, the following problems occur.
Case 1) When recommending content is determined from the user's preference information 1. Since the user asked a questionnaire to generate the preference information, if the user erroneously inputs the questionnaire answer, the user's hobby is not necessarily required. And the recommended content that matches the user's preference is not always transmitted, and the user's hobbies and preferences may not be reflected. That is, the reliability of the information about the hobbies and preferences may be low. 2. After a questionnaire, after a while, the user's hobbies and preferences may change, and the reliability of the hobbies and preferences may decrease. In addition, according to changes in the hobbies and preferences, Repeating the questionnaire is troublesome for the user.
3. The service provider needs to know in advance that the user whose preference information is X tends to purchase the content 1 in order to transmit the recommended content effectively. For example, a user who has set soccer as preference information needs to know that he tends to purchase soccer tickets.
[0004]
Case 2) When the recommended content is determined from the content purchased by the user without depending on the preference information in order to transmit the recommended content effectively 1, the person who purchases the content 1 tends to purchase the content 2 as well. Need to know in advance. For example, it is necessary to assume that a user who purchases a soccer ticket tends to purchase soccer support goods.
Case 3) When deciding a user who is a target of recommending a content belonging to a certain category 1, use a user data reference function to extract a user whose value of preference information X of the user matches a search condition such as A or more. In this case, the service provider needs to know that the user who purchases the content belonging to a certain category has a preference information value of X or more.
[0005]
The present invention has been made in view of such circumstances, and an object of the present invention is to provide a recommendation device that can enhance the reliability of preference information of a user who is a target for recommending content.
Another object of the present invention is to provide a recommendation device capable of improving the effect of an advertisement based on content purchased by a user.
[0006]
[Means for Solving the Problems]
In order to achieve the above object, the present invention provides a user action history database that stores a history of content transmitted to a user terminal as history information in association with date and time information at which the content was transmitted, and stored in the action history database. A correlation analysis unit that analyzes a plurality of contents transmitted within a predetermined period as correlated contents based on the obtained history information, an analysis result database storing analysis results of the correlation analysis unit, and a user terminal. And a distribution unit that, when a content distribution request is issued from the analysis unit, reads content correlated with the requested content from the analysis result database, and transmits an advertisement of the read correlated content to the user terminal. It is characterized by the following.
[0007]
Further, the present invention correlates a plurality of contents transmitted within a predetermined period based on history information stored in association with the history of the content transmitted to the user terminal and the date and time information at which the content was transmitted. Analyze as a content and generate an analysis result, and when there is a content distribution request from the user terminal, extract the content having a correlation with the requested content from the analysis result, and find the extracted correlation. A content advertisement is transmitted to the user terminal.
[0008]
In addition, the present invention correlates a plurality of contents transmitted within a predetermined period based on history information stored in association with the history of the content transmitted to the user terminal and the date and time information at which the content was transmitted. Analyzing the content as a certain content and generating an analysis result; and when there is a content distribution request from the user terminal, extracting content having a correlation with the requested content from the analysis result, and extracting the extracted correlation. Transmitting an advertisement of a certain content to the user terminal.
[0009]
In addition, the present invention correlates a plurality of contents transmitted within a predetermined period based on history information stored in association with the history of the content transmitted to the user terminal and the date and time information at which the content was transmitted. Analyzing the content as a certain content and generating an analysis result; and when there is a content distribution request from the user terminal, extracting the content having a correlation with the requested content from the analysis result, and extracting the extracted correlation. And transmitting an advertisement of a certain content to the user terminal.
[0010]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, a recommendation device according to an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a schematic block diagram showing a configuration of an advertisement distribution system to which a recommendation device according to an embodiment of the present invention is applied. In this figure, a recommendation device 1 is connected to a plurality of user terminals 2 via a network 2 and performs a service of transmitting content in response to a content transmission request from the user terminal 2 and collecting compensation from the user. Specifically, the content is advertised from the recommendation device 1 to the user terminal 2, and the content is sold in response to a transmission request as a purchase instruction from the user terminal 2. The recommendation device 1 will be described with reference to FIG. FIG. 2 is a schematic block diagram illustrating a configuration of the recommendation device 1 according to the first embodiment.
[0011]
In FIG. 2, the user action history database 11 stores a history of contents transmitted to the user terminal 2 and a category set for separating the contents in association with each other.
Based on the information stored in the action history database 11, the classification analysis unit 12 responds to the number of groups specified in advance so that distribution histories with a high degree of similarity in distribution of the number of distributions of the category including the distributed content are collected. Classification analysis into multiple classification groups. In addition, the classification analysis unit 12 totalizes data for performing classification analysis.
The distribution count average value calculation unit 13 calculates the average value of the distribution count in the category based on the analysis result of the classification analysis unit.
[0012]
The ratio calculation unit 14 calculates, for each category, the ratio between the average value of the number of distribution times of the category in the classification group and the average value of the distribution times of all the categories in the classification group based on the calculation result of the distribution number average value calculation unit. It is calculated every time.
The similarity calculating unit 15 classifies users into classification groups, and calculates the similarity (hereinafter, referred to as distance from the center of gravity) between the average distribution count of each category of the classified group and the distribution history of the content to the user. I do.
The priority calculation unit 16 determines a priority (hereinafter, referred to as a coefficient P) for each category in the classification group based on the approximation calculated by the approximation calculation unit 15 and the ratio calculated by the ratio calculation unit 14. Is calculated.
The preference value calculation unit 17 calculates a preference value based on the coefficient P calculated by the priority calculation unit 16 and a preset coefficient.
[0013]
The user database 18 stores the preference value calculated by the preference value calculation unit 17 and the category corresponding to the preference value in association with each other as user preference information. In this preference information, for example, as shown in FIG. 6 (a), the feature categories are set in the descending order of preference from 1 to n (n is a natural number of 2 or more) and have the same classification group (classification number). Are all set with the same feature category. The preference value set in this category indicates the degree of preference of each user in the category, and indicates that the larger the preference value, the stronger the degree of preference.
[0014]
The distribution unit 19 distributes a content advertisement based on the preference information in the user database 18.
The content database 20 stores the content to be transmitted to the user terminal 2. As shown in FIG. 3, the contents are classified into categories, and the categories are stored in association with preference information. At least one category is stored in association with this preference information. In addition, the content database 20 stores advertisement information serving as an advertisement for the content.
Each time content is transmitted to the user terminal 2, the access management unit 21 stores which content is transmitted to which user in the action history database 11.
[0015]
Next, the operation of the recommendation device 1 having the above-described configuration will be described with reference to FIGS. First, the access management unit 21 of the recommendation device 1 sequentially stores the history of the content transmitted to the user terminal 2 in the user behavior history database 11 (Step S11 in FIG. 4). When the distribution history of the content is stored in the action history database 11, the classification analysis unit 12 performs a tallying process for performing the classification analysis. That is, the number of distributions for each category into which the content transmitted to the user terminal 2 is classified is counted (step S12 in FIG. 4). Then, the process of step S12 is executed for each user (step S13 in FIG. 4). When the process of step S13 is completed, the classification analysis unit 12 classifies the distribution history of users having a high degree of similarity in distribution of the number of distributions of each category so as to gather (step S14 in FIG. 4).
[0016]
After being classified and analyzed by the classification analysis unit 12, the data is classified into the number of classification groups (for example, five groups) according to the number of classification groups specified in advance. After being classified into the classification group, the distribution count average value calculation unit 13 calculates an average value based on the distribution count of the category including the distributed content and the number of users belonging to the same classification number. As a result of this calculation, for example, as indicated by reference numeral A in FIG. 5A, the average value of the number of distribution requests for the content included in the category of “soccer” of the user belonging to the classification number 1 is 82 times. It is shown.
[0017]
When the average value is calculated by the distribution count average value calculation unit 13, the ratio calculation unit 14 calculates the average value of the distribution count of the category in the classification group and the total distribution count in the classification group based on the calculation result of the distribution count average value calculation unit. A ratio R to an average value obtained by averaging the number of distribution times of the category is calculated for each category using the following equation (1).
(Average of the number of times of the feature category within the category) / (Average of the number of times of the feature category in all the categories) ... (1)
Here, the larger the value of the ratio R, the stronger the preference. When the ratio R calculated by the equation (1) is multiplied by 100 and expressed as a percentage, the result is shown in FIG. 5B. Here, as one example, the ratio of the average value of the number of purchases of the content associated with the soccer of the remaining user to the average value (82 times) of the users belonging to the category 1 is calculated, and expressed as a percentage, it is 230%. (FIG. 5, symbol (B)).
[0018]
In FIG. 5B, when it is 100% or more, it indicates that the content distribution request for this category is higher than the overall average. If the ratio R> set value A (at least 1 or more; the average value is not determined to be strong), the user's preference is high. These categories are rearranged in ascending order of the value of the ratio R, and are temporarily stored in the user database 18 as feature categories. The setting value A is determined by the administrator of the recommendation device 1.
[0019]
When the ratio R is calculated by the ratio calculation unit 14, the approximation degree calculation unit 15 calculates the distance from the center of gravity between the distribution count average value of each category of each user classified into the classification group and the distribution history of the content to the user. Calculate. In the calculation result of the distance from the center of gravity, if the user belonging to a certain classification group completely matches the pattern of the average value of the category of the classification group, the distance from the center of gravity is 0, and the value of the distance from the center of gravity is zero. The smaller the value, the higher the commonality with the pattern of the average value of the category of the classification group. The result of calculating the distance from the reception is shown as in FIG.
[0020]
Next, when the distance from the center of gravity is calculated, the priority calculation unit 16 classifies the classification based on the distance from the center of gravity calculated by the approximation degree calculation unit 15 and the ratio R calculated by the ratio calculation unit 14. The coefficient P is calculated for each category in the group using (2) shown below.
Coefficient P = (Average number of actual times of feature category in that category) ÷ (Average actual number of times of feature category in all categories) × (1-Distance from center of gravity) (2)
[0021]
When the coefficient P is calculated, the preference value calculation unit 17 calculates the coefficient P using the following equation (3).
Preference value = ((Max−Min) × coefficient P + Min × Rate−Max) ÷ (Rate−1) (3)
Here, Rate is the ratio, Max is the upper limit of the preference value, and Min is the lower limit of the preference value, and these are set in advance by the administrator of the recommendation device 1.
[0022]
Here, when Rate = 3 is set by the administrator of the recommendation apparatus 1, all the users whose ratio R is 3 or more are determined to have the preference value as the value of Max. When the preference value is a negative value in Expression (3), the value of Min is set as the preference value. Further, a user having an average preference has a value equal to Min.
[0023]
When the preference value is calculated, the preference value calculation unit 17 stores the calculated preference value and the feature category temporarily stored in the user database 18 in the user database 18 in association with each other. The stored preference information is, for example, information as shown in FIG.
[0024]
Then, when there is a content distribution request from the user terminal 2, the preference information corresponding to the user terminal 2 that transmitted the content distribution request is read from the user database 18 with reference to the user database 18, and is included in the preference information. Based on the category and the preference value, the content and the advertisement information of the content are distributed to the user terminal 2.
[0025]
Next, the case of calculating the above-described preference value will be further described.
RATE = 3, Max = 10, Min = 0, and the degree of music liking of the user “Suzuki” is calculated as follows. Since the classification group of “Suzuki” is “2”, the category “music” in this classification group “2” has a ratio R of 1.7 from FIG.
Coefficient P = 1.7 × (1-0.2) = 1.36
Preference value = ((10-0) × 1.36 + 0 × 3-10) / (3-1) = 1.8
It becomes. This preference value is calculated for each user classification and set as preference information. Therefore, users of the same classification group have the same feature category, but have different preference values because the distance from the center of gravity differs for each user. As a result, the degree of the hobby or preference of each user can be reflected in the feature category.
[0026]
In the embodiment described above, the distribution unit 19 may distribute an advertisement of a content having no history distributed to the user terminal 2. The operation of the recommendation device 1 in this case will be described with reference to the flowchart of FIG. That is, when distributing an advertisement to the user terminal 2, the distribution unit 19 reads from the user database 18 the preference information of the user to whom the advertisement is distributed (step S <b> 51), and records the history of distribution of the content to the user Is read from the user action history database 11 (step S52). Then, of the contents included in the category set in the user's preference information, contents having no history distributed to the user terminal 2 are extracted (step S53), and the advertisement information of the extracted contents is read from the contents database and The data is distributed to the terminal 2 (step S54).
[0027]
For example, for a user whose preference information is set to “music liking degree” and who already has a history of purchasing “classical CD” content, among advertisements of “classical CD” and “concert ticket”, Only "concert ticket" ads are delivered. As a result, it is possible to distribute only advertisements related to contents to be sold in the future without distributing advertisements related to already distributed contents, thereby reducing unnecessary use of advertisements regarding already sold contents and improving the efficiency of content advertisement. Can be.
[0028]
Next, the recommendation device 1 according to the second embodiment will be described. FIG. 8 is a schematic block diagram for describing a configuration of the recommendation device 1 according to the second embodiment. In this figure, parts corresponding to the respective parts in FIG. 2 are denoted by the same reference numerals, and description thereof will be omitted. The user action history database 11 stores the history of the content transmitted to the user terminal 2 as history information in association with the date and time information at which the content was transmitted. The correlation analysis unit 30 performs a correlation analysis that associates a plurality of contents used within a predetermined period as correlated contents based on the history information stored in the action history database 11. Here, the predetermined period is a period determined by the administrator, for example, within 5 minutes, within 30 minutes, within 1 hour, and the like. Is a period that is set to determine whether or not they have been purchased together. This correlation analysis is performed at every analysis timing designated in advance by the administrator, for example, every day or every month.
[0029]
The analysis result database 31 stores the analysis result of the correlation analysis unit 30. When there is a content distribution request from the user terminal 2, the distribution unit 19 reads content having a correlation with the requested content from the analysis result database, and transmits the read advertisement of the content having the correlation to the user terminal 2. I do.
[0030]
Next, the operation of the recommendation device 1 shown in FIG. 8 will be described with reference to FIGS. Here, a description will be given of a case where a correlation analysis is performed at an analysis timing every day as content having a correlation when two or more contents are purchased without an interval of 30 minutes or more.
Upon reaching the analysis timing, the correlation analysis unit 30 totals the history information stored in the action history database 11 (Step S61 in FIG. 9), and, from the totaled history information, two or more contents within 30 minutes. Is detected, and information of two or more contents purchased within 30 minutes is generated as a record (step S62 in FIG. 9). This record generation processing is performed for each user, and the records are totaled (step S63 in FIG. 9).
[0031]
Next, the correlation analysis unit 30 performs a correlation analysis on the content in the record for each record, and calculates the degree of correlation (step S64 in FIG. 10). Here, for example, a calculation result as shown in FIG. 10A is output (Step S65 in FIG. 10). Then, the correlation analysis unit 30 extracts, for each content, a combination of the content and a content whose degree of correlation is equal to or greater than a threshold value specified in advance by an administrator, and stores the extraction result in the analysis result database 31 as analysis result data ( Step S66).
[0032]
Next, when there is a request for content distribution from the user terminal 2, the distribution unit 19 reads from the analysis result database 31 the content that is correlated with the content requested to be distributed, and reads the advertisement of the read correlated content. The information is read from the content database 20, and the read advertisement information is transmitted to the user terminal 2 together with the requested content.
[0033]
For example, when the distribution request of the content B is made from the user terminal 2, the content C of the correlated content 1 and the content D of the correlated content 2 corresponding to the content B are determined based on the analysis result data shown in FIG. It is transmitted to the user terminal 2 as advertisement information.
[0034]
According to the embodiment described above, since the advertisement of the content to be delivered to the user is determined by performing the correlation analysis, the reliability is high, and the trouble of having the user set the preference information is omitted. Advertising can be efficiently distributed.
[0035]
Next, a third embodiment will be described. FIG. 11 is a schematic block diagram illustrating the configuration of the recommendation device 1 according to the third embodiment. In this figure, parts corresponding to the respective parts in FIG. In this figure, a user database 11 stores user attribute information. This attribute information includes, for example, age, gender, annual income, etc., and indicates individual attributes.
[0036]
The user attribute condition generation unit 40 discriminates and analyzes the attribute of the user who transmitted the content based on the history information stored in the user action history database 11 and the attribute information stored in the user database 18, and executes the advertisement of the content. Generate user attribute conditions to be distributed. The distribution unit 19 distributes a content advertisement based on the user attribute condition generated by the user attribute condition generation unit 40.
[0037]
Next, the operation of the recommendation device 1 in this embodiment will be described with reference to FIGS. The user attribute condition generation unit 40 reads the user's action history from the user action history database 11 (step S71 in FIG. 12), totals the number of distributions for each category of the read action history, and outputs user attribute information from the user database 18. The read attribute information is associated with the action history (step S73 in FIG. 12). Then, the processing from step S71 in FIG. 12 to step S73 in FIG. 12 is performed for each user (step S74 in FIG. 12), and the processing result is discriminated and analyzed (step S75 in FIG. 12).
[0038]
As a result of the discriminant analysis, for example, as shown in FIG. 13, the attribute tendency of a user who has delivered a large number of results is obtained as a user attribute condition for each category, and stored in the analysis result database 31.
[0039]
Next, when there is an access from the user terminal 2, the distribution unit 19 reads the attribute information of the user of the user terminal 2 who has accessed from the user database 18, and the read attribute information matches the user attribute condition. Serve ads for content.
[0040]
Here, for example, a category A advertisement is delivered to a user of preference information in which a woman is 40 years or younger and has a preference value of 3 or more and category B is set.
[0041]
When advertising the content, the distribution unit 19 reads the user attribute condition of the category including the content from the analysis result database 31 and, from the user database 18, the user in which the preference information corresponding to the read user attribute condition is set. A plurality of data may be read and transmitted to each user terminal 2 corresponding to the user attribute condition.
[0042]
According to the above-described embodiment, the user attribute condition to which the advertisement is distributed is generated based on the history information of the distribution of the content. Therefore, the advertisement is distributed only to the users who have a high probability of purchasing the content. This makes it possible to efficiently advertise the content.
[0043]
It should be noted that among the recommendation devices 1 in the first to third embodiments in the above embodiment, the recommendation devices 1 in at least two embodiments are combined and configured as one recommendation device 1, and an advertisement distribution service is provided. May be implemented.
[0044]
Also, the functions of the classification analysis unit 12, the distribution count average value calculation unit 13, the ratio calculation unit 14, the approximation degree calculation unit 15, the priority calculation unit 16, the preference value calculation unit 17, and the distribution unit 19 in FIG. 8, a program for realizing the functions of the correlation analysis unit 30 and the distribution unit 19 or the functions of the user attribute condition generation unit 40 and the distribution unit 19 in FIG. The program recorded on the medium may be read into a computer system and executed to perform setting of user preference information and distribution processing of a content advertisement. Here, the “computer system” includes an OS and hardware such as peripheral devices.
[0045]
The “computer system” also includes a homepage providing environment (or a display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, and a CD-ROM, and a storage device such as a hard disk built in a computer system. Further, a “computer-readable recording medium” refers to a communication line for transmitting a program via a network such as the Internet or a communication line such as a telephone line, which dynamically holds the program for a short time. In this case, it is also assumed that a program that holds a program for a certain period of time, such as a volatile memory in a computer system serving as a server or a client in that case, is included. Further, the above-mentioned program may be for realizing a part of the above-mentioned functions, or may be for realizing the above-mentioned functions in combination with a program already recorded in a computer system.
[0046]
As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to the embodiments, and includes a design and the like without departing from the gist of the present invention.
[0047]
【The invention's effect】
As described above, according to the present invention, the preference information of the user is set based on the history of the content distributed to the user. As a result, it is possible to prevent the reliability of the preference information from being lowered, thereby improving the reliability of the user's preference information and improving the effect of the advertisement.
[Brief description of the drawings]
FIG. 1 is a schematic block diagram showing a configuration of an advertisement distribution system to which a recommendation device according to an embodiment of the present invention is applied.
FIG. 2 is a schematic block diagram illustrating a configuration of a recommendation device 1 according to the first embodiment.
FIG. 3 is a diagram for explaining a relationship among content, category, and preference information.
FIG. 4 is a drawing for explaining the operation of the recommendation device 1 in the first embodiment.
FIG. 5 is a diagram for explaining an operation of the recommendation device 1 according to the first embodiment.
FIG. 6 is a diagram for explaining an operation of the recommendation device 1 according to the first embodiment.
FIG. 7 is a flowchart for explaining the operation of the recommendation device 1 in the first embodiment.
FIG. 8 is a schematic block diagram illustrating a configuration of a recommendation device 1 according to a second embodiment.
FIG. 9 is a diagram for explaining the operation of the recommendation device 1 according to the second embodiment.
FIG. 10 is a diagram for explaining an operation of the recommendation device 1 according to the second embodiment.
FIG. 11 is a schematic block diagram illustrating a configuration of a recommendation device 1 according to a third embodiment.
FIG. 12 is a diagram for explaining the operation of the recommendation device 1 according to the third embodiment.
FIG. 13 is a diagram for explaining the operation of the recommendation device 1 according to the third embodiment.
[Explanation of symbols]
REFERENCE SIGNS LIST 1 recommendation device 2 user terminal 11 action history database 12 classification analysis unit 13 distribution count average value calculation unit 14 ratio calculation unit 15 approximation degree calculation unit 16 priority calculation unit 17 preference value calculation unit 18 user database 19 distribution unit 20 content database 21 access management unit 30 correlation analysis unit 31 analysis result database 40 user attribute condition generation unit

Claims (4)

A user action history database that stores a history of content transmitted to the user terminal as history information in association with date and time information when the content is transmitted,
Based on history information stored in the behavior history database, a correlation analysis unit that analyzes a plurality of contents transmitted within a predetermined period as correlated contents,
An analysis result database storing analysis results of the correlation analysis unit;
A distribution unit that, when a content distribution request is issued from the user terminal, reads content correlated with the requested content from the analysis result database, and transmits an advertisement of the read correlated content to the user terminal.
A recommendation device comprising: Analyzing a plurality of contents transmitted within a predetermined period as correlated contents based on history information stored in association with the history of the contents transmitted to the user terminal and the date and time information at which the contents were transmitted. To generate analysis results,
When there is a content distribution request from a user terminal, extracting content having a correlation with the requested content from the analysis result, and transmitting an advertisement of the extracted correlated content to the user terminal. Ad delivery method to be. Analyzing a plurality of contents transmitted within a predetermined period as correlated contents based on history information stored in association with the history of the contents transmitted to the user terminal and the date and time information at which the contents were transmitted. Generating an analysis result by using
Extracting a content having a correlation with the requested content from the analysis result, and transmitting an advertisement of the extracted correlated content to the user terminal, when a content distribution request is issued from the user terminal. An advertisement distribution program to be executed by a computer. Analyzing a plurality of contents transmitted within a predetermined period as correlated contents based on history information stored in association with the history of the contents transmitted to the user terminal and the date and time information at which the contents were transmitted. Generating an analysis result by using
Extracting a content having a correlation with the requested content from the analysis result, and transmitting an advertisement of the extracted correlated content to the user terminal, when a content distribution request is issued from the user terminal. A computer-readable recording medium on which an advertisement distribution program to be executed by a computer is recorded.

Images