KR20190044592A - Dynamic multi-dimensional analysis based advertisements recommendation apparatus and method - Google Patents

Abstract

The present invention relates to an advertisement recommending device to recommend an advertisement suitable for a user. According to the present invention, the advertisement recommending device comprises a memory to store user tendency and a processor electrically connected to the memory. The processor includes: a function (a) of checking occurrence of time serial transition about the user tendency in accordance with selection for at least one of content and an advertisement; a function (b) of updating the user tendency realized in a user tendency set sorted by the latest selected user tendency when the time serial transition of the user tendency occurs; a function (c) of performing dynamic multi-dimensional analysis for generating a preferred content population and a preferred advertisement population based on the updated user tendency; and a function (d) of providing at least one advertisement included in the preferred advertisement population in a process of viewing predetermined content even if the predetermined content not included in the preferred content population is selected.

Description

[0001] DYNAMIC MULTI-DIMENSIONAL ANALYSIS BASED ADVERTISEMENTS RECOMMENDATION APPARATUS AND METHOD [0002]

The present invention relates to an advertisement recommendation technique, and more particularly, to a dynamic recommendation apparatus based on a multidimensional analysis that recommends an advertisement suitable for a user who uses content.

The advertisement recommendation technology is being developed for the purpose of recommending an advertisement that can concentrate the attention of the user in the process of watching the contents. Ad recommendation technology has been developed to reflect not only the needs of the users including the user preference but also the needs of the advertisers according to the advertisement market logic. An advertiser may prefer to advertise a certain range of users in order to efficiently advertise their own products. The advertisement recommendation technique should be able to recommend the advertisement independently considering the needs of the user to recommend the optimized advertisement to the user.

Korean Patent Registration No. 10-1254372 (Apr. 31, 2014) relates to advertising recommendation based on social solidarity, and more particularly, to a specific user of network-based services providing opportunities for monetization provided to a specific user via a computer network Provided or monetization opportunity, and monetizing the opportunity for monetization may include clicking on the ad to cause the money to be made as a result, and based on the determination A monetization opportunity may be recommended to at least one other user other than that particular user of network-based services for which that particular user has social co-operation over the computer network.

Korean Patent Registration No. 10-1110755 (Jan. 20, 2012) relates to a second stage advertisement recommendation, more specifically, advertisements are recommended for insertion into audio and / or video programs, The preference score indicates how much the user likes each program and the advertisement classifier uses the advertiser's knowledge to provide a correlation factor indicating the effectiveness of the advertisement for the program, A validity measure may be obtained for each advertisement that indicates the effectiveness of the advertisement for a particular user by summing the products of the preference scores and the correlation factors.

Korean Patent Registration No. 10-1254372 (Apr. Korean Patent Registration No. 10-1110755 (Jan. 20, 2012)

An embodiment of the present invention is to provide a dynamic multi-dimensional analysis-based advertisement recommendation apparatus that recommends an advertisement suitable for a user who uses content.

An embodiment of the present invention is to provide a dynamic multi-dimensional analysis-based advertisement recommendation apparatus capable of independently analyzing user preference information, content information, and advertisement information to recommend an advertisement suitable for an individual.

An embodiment of the present invention is to provide a dynamic multi-dimensional analysis-based advertisement recommendation apparatus capable of recommending an advertisement suitable for an individual using an updated user preference set according to the occurrence of a time series transition related to user preference.

Among embodiments, a dynamic multidimensional analysis-based ad recommendation device includes a memory for storing user preferences and a processor in electrical communication with the memory, the processor comprising: (a) a selection of at least one of the content and the advertisement (B) a user preference, which is implemented as a user preference set arranged in the order of the most recently selected user preferences when a time series transition of the user preferences occurs, (C) performing a dynamic multidimensional analysis to generate a preference content population and a preference advertisement population through the updated user preference; and (d) The at least one optical < RTI ID = 0.0 > Including the ability to provide.

The user preference may be implemented as a user preference set including at least one of a selection count, a viewing time, and a number of departures in association with metadata of a selected content or advertisement when a selection regarding at least one of the content and the advertisement occurs.

The processor may further include a function of (e) re-updating the user tendency based on at least one of whether or not to select the at least one advertisement, the viewing time and the leaving time.

The function (b) includes a function of increasing the number of times of selecting the most recently selected user preference by 1 and updating the user tendency by assigning a higher ranking than the previous user preference when a time series transition is made regarding the user preference can do.

(C) a function of (c1) selecting a first user to determine a content preference and generating a content population based on the content preference, (c2) selecting a second user to determine an advertisement preference, (C3) generating a preference content population and a preference advertisement population by applying the user preference to the content population and the advertisement population.

The function (c2) may include, in the advertisement population, an advertisement determined according to the advertisement operation policy regardless of the advertisement selection of the second users.

The function (c3) includes the steps of: (c31) selecting only the content related to the user preference set among the contents in the content population to generate the preference content population; and (c32) And generating the preference advertisement population.

The function (c3) may generate the preference content population and the preference advertisement population by applying the user preference set including the selection order and the selection frequency information to the content population and the advertisement population.

Among embodiments, a dynamic multidimensional analysis-based advertisement recommendation method is performed in a dynamic multidimensional analysis-based advertisement recommendation apparatus that includes a memory for storing user propensity and a processor electrically connected to the memory. The dynamic multi-dimensional analysis-based advertisement recommendation method includes the steps of: (a) confirming occurrence of a time-series transition related to a user preference in accordance with occurrence of selection regarding at least one of contents and advertisements; (b) Updating a user preference, which is implemented with a user preference set sorted in order of the most recently selected user preference, if a transition occurs; (c) updating a preference content population and a preference advertisement population Performing a multi-dimensional analysis; and (d) providing at least one advertisement in the preference advertisement population during a viewing process of the specific content even if specific content not selected in the preference content population is selected.

The disclosed technique may have the following effects. It is to be understood, however, that the scope of the disclosed technology is not to be construed as limited thereby, as it is not meant to imply that a particular embodiment should include all of the following effects or only the following effects.

The dynamic multi-dimensional analysis-based advertisement recommendation apparatus according to an embodiment of the present invention can independently recommend user's recommendation information by analyzing user preference information, content information, and advertisement information independently.

The dynamic multi-dimensional analysis-based advertisement recommendation apparatus according to an exemplary embodiment of the present invention can recommend an advertisement suitable for an individual using an updated user preference set according to occurrence of a time series transition related to user preference.

1 is a view for explaining an advertisement recommendation system according to an embodiment of the present invention.
2 is a block diagram illustrating the dynamic multi-dimensional analysis-based advertisement recommendation apparatus of FIG.
Figure 3 is a block diagram illustrating the functional elements of the processor in Figure 2;
4 is a block diagram showing a module element of the multidimensional analysis performing unit in FIG.
5 is a flowchart illustrating a dynamic multi-dimensional analysis-based advertisement recommendation procedure performed by the processor in FIG.
6 is a flowchart illustrating a dynamic multidimensional analysis process performed by the multidimensional analysis performing unit shown in FIG.

The description of the present invention is merely an example for structural or functional explanation, and the scope of the present invention should not be construed as being limited by the embodiments described in the text. That is, the embodiments are to be construed as being variously embodied and having various forms, so that the scope of the present invention should be understood to include equivalents capable of realizing technical ideas. Also, the purpose or effect of the present invention should not be construed as limiting the scope of the present invention, since it does not mean that a specific embodiment should include all or only such effect.

Meanwhile, the meaning of the terms described in the present application should be understood as follows.

The terms " first ", " second ", and the like are intended to distinguish one element from another, and the scope of the right should not be limited by these terms. For example, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

It is to be understood that when an element is referred to as being "connected" to another element, it may be directly connected to the other element, but there may be other elements in between. On the other hand, when an element is referred to as being "directly connected" to another element, it should be understood that there are no other elements in between. On the other hand, other expressions that describe the relationship between components, such as "between" and "between" or "neighboring to" and "directly adjacent to" should be interpreted as well.

It is to be understood that the singular " include " or " have " are to be construed as including the stated feature, number, step, operation, It is to be understood that the combination is intended to specify that it does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

In each step, the identification code (e.g., a, b, c, etc.) is used for convenience of explanation, the identification code does not describe the order of each step, Unless otherwise stated, it may occur differently from the stated order. That is, each step may occur in the same order as described, may be performed substantially concurrently, or may be performed in reverse order.

The present invention can be embodied as computer-readable code on a computer-readable recording medium, and the computer-readable recording medium includes all kinds of recording devices for storing data that can be read by a computer system . Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Commonly used predefined terms should be interpreted to be consistent with the meanings in the context of the related art and can not be interpreted as having ideal or overly formal meaning unless explicitly defined in the present application.

The advertisement recommendation technology recommends an advertisement suitable for the user as an advertisement. The advertisement recommendation technique not only analyzes the user's preference but also analyzes the content including the video used by the user and the analysis of the advertisement provided to the user, It is possible to determine an appropriate advertisement. Multidimensional analysis implies that the factors affecting the recommendation of advertisements in the user side, content side, and advertisement side are analyzed and used independently.

Dynamic multidimensional analysis implies that, using three independent analysis results, the final advertisement provided to the user is determined by reflecting the dynamically changing user tendency. More specifically, the dynamic multidimensional analysis may be based on user preferences generated by dynamically reflecting the results of time series transitions on user preferences resulting from the selection of at least one of the user's content and / And generating a population.

1 is a view for explaining an advertisement recommendation system according to an embodiment of the present invention.

Referring to FIG. 1, the advertisement recommendation system 100 includes a user terminal 110, an advertisement recommendation device 130, and a database 150.

The user terminal 110 corresponds to a computing device that can use various contents including a movie and a drama, and may be implemented by a smart phone, a notebook computer, or a computer. . ≪ / RTI > The user terminal 110 may be connected to the advertisement recommendation device 130 via a network and at least one user terminal 110 including the user terminal 1 to the user terminal n may be concurrently connected to the advertisement recommendation device 130 .

For example, if the user terminal 110 is a smart phone, the user can view the video through the smart phone. A user can download a movie file such as a movie and a drama to a smart phone, and can watch the corresponding movie through a movie playing APP. In addition, a user can watch moving pictures such as movies and dramas in real time through a web browser installed on a smart phone or a video sharing APP.

The advertisement recommendation apparatus 130 may be implemented as a computer or a server corresponding to a program that provides recommended advertisements to the user terminal 110 based on user preference or the like through a network. The advertisement recommendation apparatus 130 may be wirelessly connected to the user terminal 110 via Bluetooth, WiFi, or the like, and may exchange data with the user terminal 110 via the network.

The advertisement recommendation apparatus 130 may be embodied including a database 150 and may be implemented separately from the database 150. When implemented separately from the database 150, the advertisement recommendation apparatus 130 may be connected to the database 150 to transmit and receive data.

The database 150 may store various information that the advertisement recommending device 130 uses to recommend an advertisement suitable for an individual. For example, the database 150 may store user preference information, content analysis information, and advertisement analysis information.

The user preference information may include user preference information composed of dynamic information such as user profile information composed of static information such as personal information of the user and content usage information and advertisement viewing information collected from the user terminal 110 .

The user preference information may correspond to criteria for selecting content and advertisement for each user. More specifically, the user preference information may include user content preference information that provides a content selection criterion and user preference information that provides an advertisement selection criterion. For example, when the content preference genre included in the content preference information of a specific user corresponds to 'horror', the user may select a category of 'horror' genre among a plurality of genre contents, It may be higher than the probability of selecting the content to which it belongs. Even if the content that can be selected by the user belongs mostly to the 'sports' genre, the user may have the highest probability of selecting the content belonging to the 'horror' genre when there is content belonging to the 'horror' genre.

In addition, if the advertisement preference genre included in the advertisement preference information of a specific user is 'sports', if an advertisement of the 'sports' genre is provided in the middle even if the user views the content of the 'horror' genre preferred by the user, The probability of stopping the viewing and viewing the advertisement may be higher than the probability of viewing the advertisement of the other genre.

The user preference information may include user preference information implemented as a user preference set. The user preference set may correspond to an ordered user preference list. The user preference set can be generated by including each of the user preferences before and after transition as an element when the user's preference changes in a time-wise manner by the selection of the user's content or advertisement. Here, the time series transition may correspond to a change in preference with the passage of time.

In one embodiment, the user preference is implemented as a user preference set that includes at least one of the number of selections, the viewing time, and the number of times of exit in association with the metadata of the selected content or advertisement when a selection of at least one of the content and the advertisement occurs . More specifically, the user preference set may be implemented with elements including at least one item included in the content or metadata of the advertisement, and may include at least one item included in the content or metadata of the advertisement, Time, and number of times of exit.

For example, a user preference set can be implemented by including as an element a 'sport item' belonging to 'sport' such as 'football', 'basketball', 'volleyball' , 'Genre' belonging to 'movie' such as '(melody', 5), ('comic', 3)} and the selection frequency of the genre.

The advertisement recommendation apparatus 130 may collect user profile information stored in the user terminal 110 with the user's consent or directly input the information through the user terminal 110. The advertisement recommendation apparatus 130 may construct user profile information based on the collected information and store the user profile information in the database 150.

The advertisement recommendation apparatus 130 may collect information related to the contents and advertisements selected and watched by the user through the user terminal 110 with the user's consent. For example, the advertisement recommendation device 130 collects various types of related information that can be used for advertisement recommendation such as viewing time and viewing day, including metadata of the content selected by the user and the advertisement, from the user terminal 110 And may be stored in the database 150.

The content analysis information corresponds to information collected by analyzing the content, and may include metadata information of the content. The content metadata may be collected through the information provided by the content provider with the content or by processing the results obtained through the content metadata analysis to collect relevant information. The content metadata may include not only general information constituting the content such as the title, type, genre, and reproduction time of the content but also information created according to certain rules such as the location and content of the content, information on the creator, have.

The content analysis information may include various types of information that can be used for advertisement recommendation besides content meta data such as an evaluation score of users on content and an average viewing time of users. For example, the advertisement recommendation apparatus 130 may process the evaluation scores evaluated by the users using the contents for each of the contents into statistical information and store them in the database 150. The advertisement recommendation device 130 processes the information that can be collected from the user terminal 110, such as the average viewing time of the content or the actual time or day of the week, Lt; / RTI >

The advertisement analysis information corresponds to the information collected by analyzing the advertisement, and may include the metadata information of the advertisement. The advertisement metadata can be collected through the information provided by the advertisement provider with the advertisement or by processing the result obtained through the analysis of the advertisement metadata. The advertisement metadata may include information created in accordance with certain rules such as the title, type, genre, and reproduction time of the advertisement as well as general information constituting the advertisement, such as the location and content of the advertisement, information about the creator, have.

The advertisement analysis information may include information processed in various forms that can be utilized for advertisement recommendation by analyzing the user's reaction to the advertisement. For example, the advertisement recommendation device 130 collects information such as the number of clicks of the actual advertisement, the time of viewing the advertisement, Lt; / RTI > In addition, the advertisement recommendation device 130 may collect information related to the content that the user was viewing at the time of providing the advertisement from the user terminal 110 and store the collected information in the database 150.

The database 150 may include at least one independent sub-database that stores information belonging to a specific range, and at least one independent sub-database may be configured as an integrated database integrated into one. In the case of at least one independent sub-database, each of the sub-databases may be wirelessly connected via Bluetooth, WiFi, or the like and may exchange data with each other via the network. In the case of an integrated database, it may include a control unit for integrating each of the sub-databases into one and managing data exchange and control flow between the sub-databases.

2 is a block diagram illustrating the dynamic multi-dimensional analysis-based advertisement recommendation apparatus of FIG.

Referring to FIG. 2, the dynamic multi-dimensional analysis-based advertisement recommendation apparatus 130 may include a processor 210, a memory 220, a user input / output unit 230, and a network input / output unit 240.

Processor 210 may execute the dynamic multidimensional analysis based advertisement recommendation procedure of FIG. 5 and may manage memory 220 that is read or created throughout the process, and may include volatile memory And the non-volatile memory. The processor 210 may control the overall operation of the dynamic recommendation apparatus 130 based on the dynamic multidimensional analysis and may be electrically connected to the memory 220, the user input / output unit 230, and the network input / Can be controlled. The processor 210 may be implemented as a central processing unit (CPU) of a dynamic multi-dimensional analysis-based advertisement recommendation device 130.

The memory 220 is implemented as a nonvolatile memory such as a solid state disk (SSD) or a hard disk drive (HDD), and is used as an auxiliary storage device , And may include a main memory implemented with a volatile memory such as a RAM (Random Access Memory).

The user input and output unit 230 includes an environment for receiving user input and an environment for outputting specific information to a user and may be a mouse, a trackball, a touch pad, a graphic tablet, a scanner, a touch screen, An input device including an adapter such as a device, and an output device including an adapter, such as a monitor or touch screen. In one embodiment, the user input and output unit 230 may correspond to a computing device connected via a remote connection, in which case the dynamic multi-dimensional analytics based advertisement recommendation device 130 may be implemented as a server.

The network input / output unit 240 includes an environment for connecting with the user terminal 110 through a network, and may include an adapter for LAN (Local Area Network) communication, for example.

Figure 3 is a block diagram illustrating the functional elements of the processor in Figure 2;

3, the processor 210 includes a time series transition confirmation unit 310, a user orientation update unit 320, a multidimensional analysis execution unit 330, an advertisement recommendation unit 340, a user orientation update unit 350, And a control unit 360.

The time series transition confirmation unit 310 can confirm occurrence of the time series transition related to the user preference. More specifically, the time series transition confirmation unit 310 searches the database 150 for user preference information associated with the selected content or advertisement when a selection regarding at least one of the content and the advertisement occurs, Or not.

For example, when a selection of content belonging to the 'comic' genre occurs, the time-series transition confirmation unit 310 searches the database 150 for the user's preference set information {'fear', 'comic' Melody '},' fear ', which is the highest priority genre, and' comic 'selected are different genres, it can be regarded that the transition of user preference finally occurs.

The user preference update unit 320 may update the user preference information. More specifically, when the time-series transition confirmation unit 310 determines that the time-series transition of the user's preferences has occurred, the user preference update unit 320 updates the user preference set You can update your propensity.

For example, when the user preference is changed from the 'horror' genre to the 'comic' genre, the user preference update unit 320 searches the database 150 for the user preference set information of the user when the generation of the time series transition is confirmed . The user preference update unit 320 updates the user preference set with {'comic', 'fear', 'melody'} if the retrieved user preference set is sorted in order of {'fear', 'comic', ' can do.

In one embodiment, the user preference update unit 320 may increase the number of selections of the most recently selected user preferences by 1 and update the user preferences by assigning the highest ranking when a time-wise transition is made regarding the user preference .

For example, if a user preference set has a pair of genres and selection frequencies as elements {('fear', 10), ('comic', 4), (' The user preference update unit 320 can update the user preference set as {('Comic', 5), ('fear', 10), ('melody', 13) have. If the content or advertisement belonging to the 'comic' genre is selected when the user preference set is {('horror', 10), ('melody', 13), (' Can update the user preference set as {('Comic', 5), ('fear', 10), ('melody', 13)

In one embodiment, the user preference update unit 320 increases the number of selection of the most recently selected user preference by 1 and assigns a higher ranking than the previous user preference when a time-wise transition is made regarding the user preference, Can be updated.

For example, if a user preference set has a pair of genres and selection frequencies as elements {('fear', 10), ('comic', 4), (' The user preference update unit 320 can update the user preference set as {('Comic', 5), ('fear', 10), ('melody', 13) have. If the content or advertisement belonging to the 'comic' genre is selected when the user preference set is {('horror', 10), ('melody', 13), (' Can update the user preference set as {('fear', 10), ('comic', 5), ('melody', 13)

The multidimensional analysis performing unit 330 may perform dynamic multidimensional analysis to generate a preferred content population and a preference advertisement population. The multidimensional analysis performing unit 330 can generate a preference content population and a preference advertisement population by selecting content and advertisements suitable for the user using the user preference information updated by the user preference updating unit 320. [ The multidimensional analysis performing unit 330 will be described in detail in the description of FIG.

The advertisement recommendation unit 340 may provide the user terminal 110 with an advertisement included in the preference advertisement population generated by the multidimensional analysis performing unit 330. The advertisement recommending unit 340 may provide at least one advertisement in the preference advertisement population to the user terminal 110 in the course of watching the specific content even if specific content not in the preferred content population is selected.

In one embodiment, the advertisement recommender 340 may sequentially provide at least one advertisement in the preference advertisement population in the order of the reordered in the course of watching the specific content. More specifically, the advertisement recommending unit 340 can sequentially provide the advertisements belonging to the reordered preference advertisement population according to the advertisement operation policy in the reproduction process of the specific content. The advertisement recommending unit 340 may provide the advertisements in order at a specific reproduction time of the specific contents or may sequentially provide advertisements at predetermined reproduction intervals.

The advertisement operation policy may correspond to an internal rule that should be kept as a top priority when the advertisement recommendation apparatus 130 provides the recommendation advertisement to the user terminal 110. [ The ad operating policies can be manually set through the ad recommendation device 130 and stored in the database 150.

The user tendency re-updating unit 350 can re-update the tendency information of the user based on the user's reaction to the recommended advertisement in real time. In one embodiment, the user preference re-updating unit 350 may re-update the user preference based on at least one of whether or not to select at least one advertisement, the viewing time, and whether or not to leave.

For example, if the user orientation is implemented as a set of user preferences and the set of user preferences is a pair of genres and selections such as {('fear', 10), ('comic', 4), (' When a recommendation advertisement belonging to the 'comic' genre is selected or viewed, the user tendency update unit 350 updates the user preference set to {('comic', 5 ',' fear ', 10) Melody ', 13)}. The advertisement recommendation apparatus 130 may store the re-updated user preference set by the user preference re-updating unit 350 in the database 150 in association with the user.

The control unit 360 controls the overall operation of the advertisement recommendation apparatus 130 and includes a time series transition confirmation unit 310, a user orientation update unit 320, a multidimensional analysis execution unit 330, an advertisement recommendation unit 340, The tendency re-updating unit 350 can manage the control flow and the data flow.

4 is a block diagram showing a module element of the multidimensional analysis performing unit in FIG.

4, the multidimensional analysis unit 330 may include a content population generation module 410, an advertisement population generation module 430, a preference population generation module 450, and a control module 470.

The content population generation module 410 may generate a population of contents as a previous step for determining an advertisement to be provided to the user terminal 110. [ More specifically, the content population generation module 410 may determine the content preference by selecting the first users, and may generate the content population based on the content preference.

Here, the first users correspond to general users who have used specific contents, and may include only a part of the users. The content preferences of the first users may correspond to universal preferences related to specific contents and may be determined based on evaluation scores performed by the users after using the contents.

The content population generation module 410 may select the first users and generate a content population according to various criteria. For example, the content population generation module 410 may sort content based on the contents selected most by the users and generate a content population composed of contents in a specific ranking according to the ordered order. The content population generation module 410 can sort content based on the content viewed for the longest time by users and generate a content population composed of contents within a specific ranking according to the ordered order.

The ad population generation module 430 may generate a population of advertisements as a previous step for determining an advertisement to be provided to the user terminal 110. [ More specifically, the advertisement population generation module 430 may select the second users to determine the advertisement preferences, and may generate the advertisement population based on the advertisement preferences.

Here, the second users correspond to general users who have used a specific advertisement, and may include only a part of the users. The advertisement preference of the second users may correspond to the general preference related to the specific advertisement, and may be based on the number of times the user clicks on the advertisement after viewing the advertisement, the time when the advertisement is viewed, . ≪ / RTI >

The advertisement population generation module 430 may select the second users and generate advertisement populations according to various criteria. For example, the advertisement population generation module 430 may sort advertisements based on the advertisements most frequently clicked by users and generate an advertisement population composed of advertisements in a specific ranking according to the ordered order. The ad population generation module 430 may generate an ad population composed of advertisements in a specific order according to the order in which the users have viewed the advertisement for the longest time or the least dropped advertisement.

In one embodiment, the ad population generation module 430 may include advertisements determined in accordance with an advertising operating policy in an ad population, regardless of ad selection by the second users. More specifically, the advertisement population generation module 430 may generate an advertisement population by including a specific advertisement according to an advertisement operation policy reflecting the intention of the advertiser. For example, the advertisement population generation module 430 may generate an advertisement population based on the advertisement preferences of the second users, and then add the advertisement of the specific advertiser to the advertisement population according to the advertisement operation policy. The advertisement population generation module 430 can place the advertisement of the specific advertiser at the top of the advertisement population regardless of the ranking even if the advertisement population is generated according to the specific ranking and can position the advertisement by designating a specific position.

The preference population generation module 450 can generate a preference content population and a preference advertisement population by applying a user preference to the content and the advertisement population generated by the content population generation module 410 and the advertisement population generation module 430 have. More specifically, the preference population generation module 450 generates a preference content population or a preference advertisement population by extracting only the content or advertisement belonging to the user preference set that implements the user preference with respect to the content population or the advertisement population generated according to the specific ranking can do.

The preference population generation module 450 may generate the preference contents and the preference advertisement populations generated by applying the user preference when the content and the advertisement population are generated according to the specific ranking, according to the ranking. For example, in the 'movie' content population, the movie 1, movie 2, and movie 3 belonging to the 'horror' genre are ranked 5th, 14th and 35th respectively in the categories of 'fear', 'melody' ', The ranking of the movies 1, 2, and 3 in the' horror 'genre included in the preferred content population is as follows: Movie 1, Movie 2, and Movie 3 It can be ranked 1st, 2nd and 3rd respectively.

In one embodiment, the preference population generation module 450 selects only the content associated with the user preference set among the contents in the content population to generate a preference content population, selects only the advertisements associated with the user preference set among the advertisements in the advertisement population, You can create a population.

For example, if the user preference set is {'soccer', 'basketball', 'volleyball'}, then the preferred population generation module 450 selects only the content associated with the highest priority 'soccer' The preferred content population can be selected to generate the preferred content population by selecting only the advertisement related to the 'volleyball' having the lowest priority among the advertisements in the advertisement population. The preference population generation module 450 can generate a preference content population by selecting all the advertisements related to at least one of 'football', 'basketball' and 'volleyball' among the contents in the content population, To create a preferred ad population.

In one embodiment, the preference population generation module 450 may apply a user preference set including selection order and selection frequency information to the content population and the advertisement population to generate a preference content population and a preference advertisement population.

For example, the preference population generation module 450 may be configured such that the advertisement population is composed of ads related to the 'sports' advertisement, and the user preference set is {('volleyball', 2), ('basketball' Soccer ', 3)}, we selected two ads with high priority among the ads related to' volleyball 'among the ads included in the ad population, and 5 of the ads related to' basketball ' And a preference advertisement population can be created by selecting three advertisements having a high priority among the advertisements related to 'soccer'. The preference population generation module 450 may generate a preference content population in the same manner for the content population.

Also, the preference population generation module 450 can generate a preference population by considering the ratio of the number of selection times. For example, the preference population generation module 450 may determine the number of selections for each event if the user preference set is {('volleyball', 2), ('basketball', 5) The number of 'basketball' and 'soccer' related ads can be 10 and 6, respectively, to create a preferred ad population if four 'volleyball' ads are included.

The control module 470 controls the overall operation of the multidimensional analysis performing unit 330 and controls the control flow and data flow between the content population control module 410, the advertisement population control module 430 and the preference population generation module 450 Can be managed.

5 is a flowchart illustrating a dynamic multi-dimensional analysis-based advertisement recommendation procedure performed by the processor in FIG.

Referring to FIG. 5, the processor 210 may check the occurrence of a time-series transition related to user preference in response to the selection of contents and advertisements through the time-series transition confirmation unit 310 (step S510).

The processor 210 may update the user preference implemented in the user preference set sorted in order of the most recently selected user preference through the user preference updating unit 320 (step S530).

The processor 210 may perform a dynamic multidimensional analysis to generate a preference content population and a preference advertisement population through a user tendency through the multidimensional analysis performing unit 330 (step S550).

The processor 210 may provide at least one advertisement in the preferred advertisement population in the course of watching the specific content through the advertisement recommending unit 340 (step S570). In one embodiment, the advertisement recommendation unit 340 may provide a specific advertisement to the user terminal 110 according to the advertisement operation policy reflecting the advertiser's intention. For example, the advertisement recommending unit 340 can provide advertisement of a specific advertiser first regardless of the ranking of the preferred advertisement population generated by the preferred population generating module 450 included in the multidimensional analysis performing unit 330 And can provide advertisements of a specific advertiser regardless of the ranking at a specific playback point or a specific playback interval.

The processor 210 may re-update the user tendency based on at least one of whether or not to select the advertisement, the viewing time, and the desire to leave the user through the user tendency updating unit 350 (step S590).

6 is a flowchart illustrating a dynamic multidimensional analysis process performed by the multidimensional analysis performing unit shown in FIG.

The multidimensional analysis performing unit 330 may generate the content population through the content population generation module 410 (step S610). The content population generation module 410 may select the first users to determine the content preference, and may generate the content population based on the content preference.

In one embodiment, the content population generation module 410 can select users who tend to use content similar to the user through collaborative filtering as a first user, and generate a content population based on content preferences of first users .

The multidimensional analysis performing unit 330 may generate an advertisement population through the advertisement population generating module 430 (step S630). The advertisement population generation module 430 may select the second users to determine the advertisement preferences and generate the advertisement population based on the advertisement preferences.

In one embodiment, the advertisement population generation module 430 may select users who have selected or watched advertisements similar to those previously selected or watched by the user through content-based filtering, as a second user, An ad population can be created based on the advertisement preference. In another embodiment, the advertisement population generation module 430 may select users who have selected or watched advertisements based on an advertisement target similar to an advertisement target of an advertisement previously selected or viewed by the user as a second user .

The multidimensional analysis performing unit 330 may generate a preference content and a preference advertisement population by applying the user preference to the content and the advertisement population through the preference population generating module 450 (step S650).

In one embodiment, the preference population generation module 450 may include a specific advertisement according to an advertising operation policy reflecting the advertiser's intention, thereby generating a preference advertisement population. For example, the preference population generation module 450 may generate a preference population by generating an advertisement population of the advertiser according to a specific ranking and generating a preference advertisement population by applying the user tendency by the advertisement population generation module 430 It can be positioned at the top of the preference advertisement population, and a specific location can be specified to position the advertisement.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims It can be understood that

100: Ad referral system
110: User terminal
130: Ad recommendation device based on dynamic multi-dimensional analysis
150: Database
210: processor 220: memory
230: user input / output unit 240: network input / output unit
310: Time series transition confirmation unit 320: User orientation update unit
330: Multidimensional analysis performing unit 340: Advertisement recommendation unit
350: user tendency re-updating unit 360:
410: Content Population Creation Module 430: Advertisement Population Creation Module
450: preference population generation module 470: control module

Claims (9)

A memory for storing user preferences; And
And a processor electrically coupled to the memory,
The processor
(a) a function of confirming occurrence of a time-wise transition with respect to a user's preference in accordance with occurrence of selection regarding at least one of contents and advertisements;
(b) a function of updating the user preference implemented as a user preference set arranged in the order of the most recently selected user preferences when the user preference time series occurs;
(c) performing dynamic multidimensional analysis to generate a preference content population and a preference advertisement population through the updated user preference; And
(d) providing at least one advertisement in the preference advertisement population during the viewing of the specific content even if a specific content not in the preference content population is selected.
The method of claim 1,
Wherein the user preference set is implemented as a user preference set including at least one of a selection count, a viewing time, and an exit count in association with metadata of the selected content or advertisement when a selection is made regarding at least one of the content and the advertisement. Based advertisement recommendation apparatus.
2. The apparatus of claim 1, wherein the processor
(e) re-updating the user tendency based on at least one of whether or not to select at least one advertisement, the viewing time, and whether to leave the advertisement.
The method of claim 1, wherein the function (b)
And a function of updating the user tendency by increasing the number of selection of the most recently selected user preference by 1 and assigning a higher ranking than the previous user preference when a time series transition related to the user preference occurs, Based multi-dimensional analysis recommendation apparatus.
The method of claim 1, wherein the function (c)
(c1) selecting a first user to determine a content preference and generating a content population based on the content preference;
(c2) a function of selecting second users to determine advertisement preferences and generating advertisement populations based on the advertisement preferences; And
(c3) generating a preference content population and a preference advertisement population by applying the user preference to the content population and the advertisement population.
6. The method of claim 5, wherein the function (c2)
And a function of including the advertisement determined according to the advertisement operation policy in the advertisement population regardless of the advertisement selection of the second users.
6. The method of claim 5, wherein the function (c3)
(c31) a function of selecting only the content related to the user preference set among the contents in the content population to generate the preferred content population; And
(c32) generating a preference advertisement population by selecting only the advertisement related to the user preference set among the advertisements in the advertisement population.
6. The method of claim 5, wherein the function (c3)
And generating the preference content population and the preference advertisement population by applying the user preference set including the selection order and the selection frequency information to the content population and the advertisement population. Ad recommendation device.
A dynamic multi-dimensional analysis-based advertisement recommendation method performed in a dynamic multi-dimensional analysis-based advertisement recommendation apparatus including a memory for storing a user tendency and a processor electrically connected to the memory,
comprising the steps of: (a) confirming occurrence of a time-series transition with respect to user preference in accordance with occurrence of a selection regarding at least one of contents and advertisements;
(b) updating the user preference implemented with a user preference set arranged in order of the most recently selected user preferences when the user preference has time-wise transition;
(c) performing dynamic multidimensional analysis to generate a preference content population and a preference advertisement population through the updated user preference; And
(d) providing at least one advertisement in the preference advertisement population during the viewing of the specific content even if specific content not in the preference content population is selected.

Images