KR101089646B1 - Method and apparatus for content recommendation - Google Patents

Abstract

Disclosed are a method and apparatus for recommending content. The method for recommending content by the content recommending apparatus may calculate a rate of change of a content section and a rate of change of a period of time using a user's content usage history, select m pieces of large fluctuations, and extract a user who uses the corresponding m pieces of content. Calculate user usage level, select top y users with high user usage level, and select and recommend a certain number of contents that other users moved the most from the current contents of each selected user as recommended content Can be.

Content, contents, recommendations

Description

Method and apparatus for content recommendation

The present invention relates to content provision, and more particularly, to a method and an apparatus capable of recommending content suitable for a user.

Advances in technology and economic growth have accelerated the dissemination of mobile telecommunication terminals that enable telephone calls on the go. In addition to the conventional telephone call function, the mobile communication terminal manufactured and sold recently has various additional functions such as an address book function, a short message (SMS) transmission and reception function, a wireless Internet access function, and a real time message transmission and reception function.

The user can download and use contents (eg, ringtones, wallpaper images, music contents, movie contents, etc.) provided by an arbitrary content providing server using a wireless Internet access function provided in the mobile communication terminal. . The types of content currently available to users accessing the wireless Internet include image content (eg, wallpaper images, celebrity photo images, etc.), music content (eg, ringtones, MP3 music, etc.), video content ( For example, movies, dramas, etc.), game content (e.g., GoStop, etc.), real-time information content (e.g., news, stock market, sports news, etc.), as well as gradually increasing its variety It is increasing.

However, in order to use such content, the user has to pay a lot of time and communication cost to access the server providing the content and search for the desired content.

The present invention is to provide a method and apparatus that can recommend content suitable for a user.

In addition, the present invention is to provide a method and apparatus capable of recommending content having a high selection possibility by analyzing a user having a high content changing possibility.

In addition, the present invention is to provide a method and apparatus capable of recommending content centered on a user who has come to change the content.

According to an aspect of the present invention, a method for recommending content by a content recommending apparatus is provided.

According to an embodiment of the present invention, a method for recommending a content by a content recommending apparatus, the method comprising: (a) calculating a rate of change of each content section and a rate of change of a period of time by using a content usage history of a user; (b) selecting m (m is a natural number) contents having a large variation using the calculated interval change rate and the period change rate; (c) extracting a user using at least one of the selected m contents based on a current time point, and using the user by using an average user's use time of the selected content, an average user's use time of the extracted user's content, and a current content use time Calculating a degree; (d) selecting y (random natural numbers) users whose use level differs from each other; And (e) selecting and recommending q (q is a natural number of 1 or more) contents having a large variation from the current contents of each of the selected users in a recent section as recommended content. Can be.

The q contents are contents that other users have moved a lot from the current contents, and the interval change rate is a ratio of a change in the number of users using the contents from the current time point to the previous time point, and the period change rate is arbitrary at any time point. It may be a ratio for the change in the number of users using the content up to the point of time.

Before step (a), further comprising the step of storing the content usage history for each user in accordance with the user's use of the content,

The step (b) may include: calculating a content reduction rate by multiplying the interval variation rate and the period variation rate by weight for each content, respectively; And arranging according to a predetermined method by using the calculated content reduction rate, and selecting m pieces of contents having a large variation.

The weight is multiplied differently to the interval change rate and the period change rate, respectively, and the sum of each weight multiplied by the interval change rate and the period change rate may be 1, respectively.

Step (c) may include extracting a user who uses the selected content at the present time; Extracting the average content usage time and the current content usage time of each extracted user using a content average usage time table for each user; Extracting a user average use time of the selected content; And calculating a user use degree using the extracted content average use time, current content use time, and user average use time.

According to another aspect of the present invention, there is provided a content recommendation apparatus for recommending content suitable for a user.

According to an embodiment of the present invention, there is provided a content recommending apparatus for recommending a content suitable for a user, comprising: a variation rate calculating unit configured to calculate a rate of change of each content section and a rate of change of a period of time by using a content usage history of a user; The m (m is a natural number) content having a large variation is selected by using the calculated interval variation rate and the period variation rate, a user who uses at least one of the selected m contents based on the current time point is extracted, and the selected content is selected. A selection unit configured to calculate a user use degree by using a user average use time of the user, an average user use time of the extracted user, and a current content use time of the user; And a content recommendation unit for selecting and recommending q (q is a natural number of 1 or more) contents having a large variation from the current contents of each of the selected users in a recent section as recommended content. .

By providing a content recommendation method and apparatus according to an embodiment of the present invention, it is possible to recommend content suitable for a user.

In addition, the present invention has the advantage of recommending content with a high possibility of selection by analyzing a user with a high possibility of changing the content.

In addition, the present invention has the advantage that can be recommended to the content centered on the user that the content change time has arrived.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, it should be understood to include all transformations, equivalents, and substitutes included in the spirit and scope of the present invention. In the following description of the present invention, when it is determined that the detailed description of the related known technology may obscure the gist of the present invention, the detailed description thereof will be omitted.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[Description of FIGS. 1 to 5]

1 is a block diagram schematically showing a content recommendation system according to an embodiment of the present invention, FIG. 2 is a diagram illustrating a content usage status table for each time point according to an embodiment of the present invention, and FIG. FIG. 4 is a diagram illustrating a user movement status table for each section according to an embodiment of the present disclosure. FIG. 4 is a diagram illustrating each content average use time table according to an embodiment of the present invention. FIG. 5 is a diagram illustrating an embodiment of the present invention. FIG. 1 is a diagram illustrating an average content usage time table for each user.

Referring to FIG. 1, a content recommendation system according to an exemplary embodiment of the present invention includes a user terminal 110, a content recommendation device 120, and a database 130.

The user terminal 110 is a terminal that can receive various contents by accessing the content recommendation apparatus 120 through a communication network. For example, the user terminal 110 may be equally applicable to all communication devices capable of communicating, such as a mobile communication terminal, a computer, a notebook computer, and a PDA.

The content recommendation device 120 performs a function of providing various contents to each user terminal 110 through a communication network. In addition, the content recommendation device 120 may collect and analyze content usage history for each user to recommend content with high availability to each user. In the present embodiment, it will be described on the assumption that the content recommendation apparatus 120 performs a function of providing various contents in addition to a function of recommending contents by collecting a history of content usage through each user terminal 110. .

Of course, according to the implementation method may be provided with a separate device for providing the content.

For example, the content may be a sound source such as coloring, but is not limited to the sound source, and may be equally applied to multimedia content in addition to the sound source.

Hereinafter, the method of recommending the content recommendation device 120 to each user with a high possibility of using the content will be described with emphasis, which will be described in more detail with reference to the accompanying drawings below.

The database 130 stores user information (eg, user identification information, phone number, etc.) and various information about the content usage history.

For example, the database 130 may store a content usage status table for each time point, a user movement status table for each section, an average usage time table for each content, and an average usage time table for each user.

In the present embodiment, the content usage status table for each time point may map and store the number of users who use (or use and download) content for each time point. That is, by analyzing the content usage status table for each time point, the increase or decrease of the number of users who use (use, download) each content according to each time point, and the ratio according to the increase and decrease may be analyzed.

The user movement status table for each section maps and changes the change status of use of each content for each section for each content. That is, it is possible to grasp the current status of which content or which content the user moved to during the period.

The average usage time table for each content maps and stores the average of user's usage time for each content for each content. In addition, the average usage time table for each content may also be mapped and stored with the standard deviation of the average usage time for each content. Here, the use time may be any one of time, day, week, and month.

 The average usage time table for each user shows the average usage time for each user's use of each content, the standard deviation of the average usage time for each content, the usage time of the current content, and the identification number of the content currently used by the user. Map and save.

As such, the database 130 may store information on the user and a history of the user's use of the content in various forms. In the present embodiment, it is assumed that the database 130 is connected to the outside of the content recommendation device 120, but the database 130 may be implemented as one component inside the content recommendation device 120.

6-7 description

FIG. 6 is a block diagram schematically illustrating an internal configuration of a content recommending apparatus according to an exemplary embodiment of the present invention, and FIG. 7 is a flowchart illustrating a method for selecting a user with high mobility according to an exemplary embodiment of the present invention.

Referring to FIG. 6, the content recommendation apparatus 120 according to the present embodiment includes a data transceiver 610, a collector 615, a register 620, a change rate calculator 625, a selector 630, and a content. The recommendation unit 635 and the server control unit 640 are configured.

The data transceiver 610 is connected to each user terminal 110 through a communication network to perform a function of transmitting and receiving data.

The collection unit 615 performs a function of collecting a history (hereinafter, referred to as a content usage history) according to arbitrary content usage from each user terminal 110. Here, the content usage history may include various information about download, content usage change, etc. for using the content of each user terminal 110.

The registration unit 620 uses a content usage history collected through the collection unit 615, according to a predetermined method, a content usage status table for each time point, a user movement status table for each section, an average usage time table for each content, Creates or updates the average content usage time table for each user.

The change rate calculator 625 calculates the period change rate and the period change rate for each content by using the content usage status table for each time point.

For example, the change rate calculator 625 may calculate a change rate of the user for each content up to the latest time point based on the current time point by using the content use status table for each time point. That is, the change rate calculator 625 derives the number of users for each content at each time point, and calculates the rate of change of the section by using the same.

For example, the change rate calculator 625 may calculate a section change rate for each content by using Equation 1 below.

는 현재 시점 k에서 콘텐츠 i를 이용하는 사용자 수를 나타내고,

는 현재 시점 직전의 시점(k-1)에서 콘텐츠 i를 이용한 사용자 수를 나타낸다.Where i represents content, k represents the current point in time,

Represents the number of users using content i at current time k,

Denotes the number of users using the content i at a time point k-1 just before the current time point.

That is, the interval change rate may be a change rate of the number of users using the content i from the previous time point k-1 to the current time point k. Accordingly, if the interval change rate is negative, it can be seen that the user using the corresponding content is in a decreasing trend. In addition, if the interval variation rate is a positive number, it can be seen that the user using the corresponding content is increasing.

In addition, the change rate calculator 625 may calculate a change rate of the user (for example, a change rate of a period) of the content for each period.

For example, the change rate calculator 625 may calculate a period change rate for each content by using Equation 2 below.

는 현재 시점 k에서 콘텐츠 i를 이용하는 사용자 수를 나타내고,

는 시점 j에서의 콘텐츠 i를 이용하는 사용자의 수를 나타낸다.Where i represents content and j and k each represents an arbitrary point of view,

Represents the number of users using content i at current time k,

Denotes the number of users using content i at time point j.

That is, the period change rate represents a change trend of the user who uses the corresponding content i during the period from an arbitrary time point j to the current time point k. Accordingly, if the period change rate is negative, the user using the content during the corresponding period is decreasing, and if the period change rate is positive, the user using the content during the corresponding period is increasing.

In the present embodiment, the interval change rate is a trend of change of the user according to the use of each content for the latest interval, and the period variation rate is defined as a trend of change of the user according to the use of each content for a period from an arbitrary time point to the present time. .

The selector 630 selects m (arbitrary natural numbers) of large fluctuations using the interval change rate and period change rate of each content calculated by the change rate calculator 625 and moves to the selected m contents. It performs the function of selecting each likely user.

For example, the selector 630 multiplies the interval variation rate and the period variation rate for each content by a predetermined weight, and adds the interval variation rate and the period variation rate multiplied by the weight, respectively, to reduce the reduction rate for each content (hereinafter, the contents). Reduction rate). Here, the decrease rate is used as a term opposite to the increase rate, and a high decrease rate is used to mean a low increase rate. Therefore, the reduction rate may be expressed as a high reduction rate when the increase rate is low when the change rate is substantially less than zero, but also when the change rate is larger than zero.

Here, the reduction rate for each content may be calculated using Equation 3 below.

Here, w1 and w2 represent weights (or adjustment coefficients), respectively. Here, w1 + w2 = 1, w1 and w2 may be set differently.

For example, when w1 is set to “0”, in calculating the reduction rate for each content, it can be seen that the reduction rate for the content is calculated by considering only the reduction rate for each period, not considering the reduction rate for the current section. .

On the other hand, when w2 is set to "0", it can be seen that in calculating the reduction rate for each content, the reduction rate for each period is not considered and is calculated considering only the reduction rate for the current section.

Each of the weights w1 and w2 may be set differently according to the weight of each period or interval variation rate for each content.

The smaller the value of Equation 3 is, the higher the content reduction rate is, and the inverse may be taken so that a large value is derived when the content reduction rate is high.

The selector 630 may sort m contents having a high content reduction rate (or a low growth rate) by sorting according to a predetermined sorting method using the calculated reduction rate for each content. Here, the sorting method may be in descending or ascending order.

In addition, the selector 630 may select a user having a high possibility of movement among the current users of the selected m pieces of content.

Hereinafter, for convenience of understanding and explanation, a method of selecting a user having a high possibility of moving with respect to m pieces of selected content will be described in detail with reference to FIG. 7.

In operation 710, the selector 630 extracts users who use m pieces of selected content at the present time.

For example, the selector 630 may extract a user who uses m pieces of selected content by using a content average use time table for each user.

In operation 715, the selector 630 extracts the average content usage time and the current content usage time of each extracted user. Here, the unit of use time may be any one of time, day, week, and month.

For example, the selector 630 may use the content average usage time (hereinafter, referred to as the extracted user content average usage time) and the current content usage time of each user extracted using the content average usage time table for each user. The current content usage time for each user can be extracted.

In operation 720, the selector 630 extracts an average use time (hereinafter, referred to as an average use time of the selected content) for each selected content.

For example, the selector 630 may extract a user average use time for each selected content using an average use time table for each content.

In operation 725, the selector 630 calculates a user usage degree using the extracted content average usage time for each user, the current content usage time for each extracted user, and the average usage time for each extracted content.

For example, the selector 630 divides user h's current content i usage time by content i's average usage time and user h's current content i usage time by user h's content average usage time. Can calculate

For example, the selector 630 may calculate a user use degree using Equation 4 below. The user usage level indicates the current usage level of the content compared to the average usage time of the content and the average usage time of the user h. The higher the value, the higher the usage level.

Here, w3 and w4 denote weights, respectively, denotes a user h's current content i usage time, ACi represents an average usage time of the content i, and AUh represents an average usage time of the user h. In addition, h represents each extracted user and i represents each of the selected contents. In addition, the sum of w3 and w4 may be 1.

Also, if w3 is 0, the user use degree is derived using the ratio of the usage time of the current content i of the user h to the average usage time of the user h. And the average usage time of the content i.

As another example, the standard deviation may be used to calculate user usage. That is, Equation 4 may be modified in such a manner that the weight is increased as the standard deviation is lower. If the standard deviation is low, it is likely to have a near-average use time, which can be reflected. That is, as the standard deviation of the content i is lower, a higher weight can be given to the usage time of the content i. In addition, the lower the standard deviation of the user h, the higher the weight of the user h can be given.

For example, Equation 4 may be modified as shown in Equation 5 by using a standard deviation.

이고,

이다.here,

ego,

to be.

In Equation 5, the maximum standard deviation of the content usage time is the maximum value among the standard deviations of the content usage time, and the maximum standard deviation of the user usage time is the maximum value of the standard deviation of the user usage time. The reason for adding 1 to each standard deviation in d1 and d2 is to prevent calculation errors from occurring because each standard deviation becomes zero. In addition, w5 and w6 are weights, respectively, and the degree of reflection of each standard deviation can be determined, and the user usage degree according to Equation 5 is determined by d1 and d2 by limiting the values of d1 and d2 within a certain range. You can prevent it. In addition, w5 and w6 can also be set differently according to a use time, respectively. For example, w5 and w6 may be set low if the usage time is less than the average usage time, and w5 and w6 may be set higher if the usage time exceeds the average usage time. In addition, the weights w5 and w6 may be applied in various ways depending on the usage time and standard deviation.

In operation 730, the sorting unit 630 sorts the calculated user use degree in a predetermined sorting method (eg, descending order, ascending order, etc.) and then selects the upper y (any natural number) users.

For example, the selector 630 may select a user whose user usage exceeds a predetermined threshold.

As another example, the selector 630 may select y users whose user use level is higher. Here, y may be preset by the administrator.

Here, the high user usage means that the user is using the content currently being used for a long time in preparation for the average usage time of the user. Therefore, a user with a higher user usage is likely to move to a new content. Can be.

Referring back to FIG. 6, the content recommendation unit 635 extracts content to be recommended for the selected user.

For example, the content recommendation unit 635 targets each of the selected users, and a predetermined number (eg, q and q is 1) of the highest rankings that other users have moved the most in the current content of each selected user. Content of the above-mentioned natural number) may be extracted as recommended content and provided to each selected user.

To this end, the section user movement status table illustrated in FIG. 3 may be used. That is, the content recommendation unit 635 extracts a predetermined number (q) of contents of the top rank that other users have moved the most in the recent section based on the selected content of each user, and selects the content to be recommended. Can be. That is, as other users who have used your current content move to other content, a certain number of the top ranked content in the most recent segment (ie, most popular with other users who have used your current content in recent segments). Content).

At this time, in the process of extracting q recommendation target content, the selected content having the highest m decrease rate may be excluded. That is, m pieces of content having a high reduction rate may not be selected as a recommendation target.

The user movement status table for each section may be calculated by the content recommendation unit 635 using the statistics stored in the database 130, and is registered by the register 620 as statistics on the user movement status for each content for each section. Of course, you can register in advance. In this case, the content recommendation unit 635 may naturally select the content to be recommended for the most recent section.

The server controller 640 calculates internal components of the content recommendation apparatus 120 according to an embodiment of the present invention (eg, the data transmitter / receiver 610, the collector 615, the registerer 620, and the variation rate calculation). Unit 625, selection unit 630, content recommendation unit 635, and the like.

8 is a flowchart illustrating a method for recommending content suitable for a user by a content recommending apparatus according to an exemplary embodiment of the present invention. Each step described below is performed by each internal component of the content recommending device 120, but will be collectively described as a content recommending device for the convenience of understanding and explanation.

In operation 810, the content recommendation device 120 calculates a section variation rate for each content by using a content usage status table for each time.

It will be described on the assumption that the content usage status table for each time point is pre-stored by the registration unit 620 based on the content usage history of the user.

For another example, the content recommendation device 120 may generate a content usage status table for each time point based on each user's content usage history, and may calculate a section variation rate for each content by using the content recommendation device 120.

In operation 815, the content recommendation device 120 calculates a period variation rate for each content by using the content usage status table for each time point.

In operation 820, the content recommendation device 120 selects m (random natural numbers) of large variations using the calculated interval change rate and period change rate of each content.

In operation 825, the content recommendation device 120 extracts each user who uses m pieces of content selected at the present time, calculates user usage for each extracted user, and then selects users with high usage. have. Since this is the same as described with reference to FIG. 7, redundant descriptions will be omitted.

In operation 830, the content recommendation device 120 selects a predetermined number of contents, which are moved most by other users, from the current contents of each selected user as the content to be recommended.

When the selected user accesses the content recommendation device 120, the content recommendation device 120 may provide the selected recommendation content to be output to the selected user.

For another example, the content recommendation device 120 uses the selected content to be selected according to the selected user by using a text message (for example, any one of a short message, a long message, and a multimedia message). It can also be provided to.

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 invention as defined in the appended claims. It will be understood that the invention may be varied and varied without departing from the scope of the invention.

1 is a block diagram schematically illustrating a content recommendation system according to an embodiment of the present invention;

2 is a diagram illustrating a content usage status table for each time point according to an embodiment of the present invention.

3 is a diagram illustrating a user movement status table for each section according to an embodiment of the present invention.

4 is a diagram illustrating each content average usage time table according to an embodiment of the present invention.

5 is a diagram illustrating a content average use time table for each user according to an embodiment of the present invention.

6 is a block diagram schematically showing an internal configuration of a content recommending apparatus according to an embodiment of the present invention.

7 is a flowchart illustrating a method of selecting a user having a high possibility of moving according to an embodiment of the present invention.

8 is a flowchart illustrating a method for recommending content suitable for a user by a content recommending apparatus according to an exemplary embodiment of the present invention.

Claims (18)

In the method for recommending content by the content recommending device, (a) calculating the rate of change of each content section and the rate of change of period of time using the user's content usage history; (b) selecting m (m is a natural number) content having a large variation using the calculated interval change rate and the period variation rate, wherein the interval change rate is based on a change in the number of users who use the content from the current time point to the previous time point. The rate of change of the period is a rate of change in the number of users using content from any point in time to any point in time; (c) extracting a user using at least one of the selected m contents based on a current time point, and using the user by using an average user's use time of the selected content, an average user's use time of the extracted user's content, and a current content use time Calculating a degree; (d) selecting y (random natural numbers) users whose use level differs from each other; And and (e) selecting and recommending q (q is a natural number of 1 or more) contents having a large variation from the current contents of each of the selected users in a recent section as recommended content. The method according to claim 1, The q content is a content recommendation method, characterized in that the content has moved a lot of other users from the current content. The method according to claim 1, Before step (a), And storing the content usage history for each user according to the user's use of the content. delete The method according to claim 1, In step (b), Calculating a content reduction rate by multiplying the interval variation rate and the period variation rate by respective weights for each content; And Selecting m contents having a large variation by arranging according to a predetermined method using the calculated content reduction rate. 6. The method of claim 5, The weight is multiplied by the interval change rate and the period change rate respectively differently, The sum of each weight multiplied by the interval change rate and the period change rate is 1, respectively. The method according to claim 1, In step (c), Extracting each user who uses the selected content at the present time; Extracting the average content usage time and the current content usage time of each extracted user using a content average usage time table for each user; Extracting a user average use time of the selected content; And And calculating a user usage degree using the extracted content average usage time, current content usage time, and user average usage time. 8. The method of claim 7, Content usage method, characterized in that the user use degree is calculated using the following equation.

Here, w3 and w4 each represent a weight,

Represents the current usage time of user i, ACi represents the average usage time of content i, AUh represents the average usage time of user h content, h represents each extracted user, and i represents the selected content. Respectively. The method according to claim 1, The step (d) Selecting the top y users from the users whose degree of use exceeds a preset threshold. In the content recommendation device for recommending content suitable for a user, A change rate calculating unit for calculating each content section change rate and period change rate using a user's content usage history, wherein the section change rate is a ratio of the change in the number of users who use the content from the current time point to the previous time point. The rate of change in the number of users using content from any point in time to any point in time; The m (m is a natural number) content having a large variation is selected by using the calculated interval variation rate and the period variation rate, a user who uses at least one of the selected m contents based on the current time point is extracted, and the selected content is selected. A sorting unit configured to calculate a user usage degree by using a user average usage time of the user, an average user usage time of the extracted user, and a current content usage time of the user; And And a content recommending unit for selecting and recommending q (q is a natural number having one or more) contents having a large fluctuation from the current content of each of the selected users in a recent section as recommended content. The method of claim 10, The q content is a content recommendation device, characterized in that the other user has moved a lot from the current content. The method of claim 10, And a registration unit for collecting and registering a content usage history according to the user's use of the content. delete The method of claim 10, The sorting unit, For each content, the interval change rate and the period change rate are multiplied and summed, respectively, to calculate a content reduction rate, and the m content with large variation is selected by sorting according to a predetermined method using the calculated content reduction rate. Characterized in that the content recommendation device. 15. The method of claim 14, The weight is multiplied by the interval change rate and the period change rate respectively differently, The sum of each weight multiplied by the interval change rate and the period change rate is 1, respectively. The method of claim 10, The sorting unit, Extracting each user using the selected content at the current time point, and extracting the average content usage time and the current content usage time of each extracted user using a content average usage time table for each user, and the user of the selected content. And extracting an average use time, and calculating a user use degree using the extracted content average use time, current content use time, and user average use time. The method of claim 16, Contents recommendation device, characterized in that the user use is calculated using the following equation.

Here, w3 and w4 each represent a weight,

Represents the current usage time of user i, ACi represents the average usage time of content i, AUh represents the average usage time of user h content, h represents each extracted user, and i represents the selected content. Respectively. The method of claim 10, The sorting unit, And selecting the top y (random natural numbers) of users from the users whose degree of use exceeds a predetermined threshold.

Images