KR20220099300A - Method, server and computer program for selecting contents - Google Patents

Abstract

A method for selecting content comprises the steps of: forming a plurality of clusters for a plurality of contents; defining a first cluster set including the plurality of formed clusters; selecting a cluster from the first cluster set based on similarity between the plurality of clusters included in the first cluster set and a second cluster set; including the selected cluster in the second cluster set; selecting a content from the selected cluster based on similarity between one or more contents pre-selected from the selected cluster and one or more contents not pre-selected from the selected cluster; and providing the selected contents to a user terminal. The selected contents correspond to any one among a plurality of preset famous contents. Therefore, a content set consisting of various contents can be selected to attract the interest of a user.

Description

Method of content selection, server and computer program

The present invention relates to a content selection method, a server and a computer program.

In general, a content recommendation system selects and recommends a plurality of contents suitable for user interests based on user information input or received into the system. However, the conventional content system has a disadvantage in that it is difficult to recommend a plurality of contents suitable for user interests when there is no previously inputted user information, that is, for a new user.

1 is an exemplary diagram for explaining a conventional content recommendation system according to an embodiment of the present invention. Referring to FIG. 1 , the conventional content recommendation system lists and presents different types of content 110 to new users. A new user may select a content suitable for interest from among the plurality of listed content 110 .

The conventional content recommendation system may utilize information on the new user based on the content selected from among the plurality of contents 110 listed by the new user.

However, the plurality of contents 110 listed in the conventional content recommendation system has a very limited disadvantage in identifying the interests of new users. Specifically, the plurality of contents 110 listed in the conventional content recommendation system are randomly selected without a separate criterion, and are not suitable for properly understanding the user's interests.

Korean Patent Publication No. 10-2096474 (Registered on March 27, 2020) Korean Patent Publication No. 10-2019-0096952 (published on August 20, 2019)

An object of the present invention is to provide a content selection method that can select a content set composed of various contents in order to understand the interests of users.

However, the technical problems to be achieved by the present embodiment are not limited to the technical problems described above, and other technical problems may exist.

As a means for achieving the above-described technical problem, an embodiment of the present invention provides a content selection method, comprising: forming a plurality of clusters for a plurality of content; defining a first cluster set including the formed plurality of clusters; selecting a cluster from the first cluster set based on a degree of similarity between a plurality of clusters and a second cluster set included in the first cluster set; including the selected cluster in the second cluster set; selecting content from the selected cluster based on a similarity between one or more contents previously selected from the selected cluster and one or more contents not previously selected from the selected cluster; and providing the selected content to a user terminal, wherein the selected content corresponds to any one of a plurality of preset famous content.

Another embodiment of the present invention provides a content selection server, comprising: a content clustering unit for forming a plurality of clusters for a plurality of content; a cluster defining unit defining a first cluster set including the formed plurality of clusters; a cluster selector configured to select a cluster from the first cluster set based on a degree of similarity between a plurality of clusters and a second cluster set included in the first cluster set; a cluster processing unit for including the selected cluster in the second cluster set; a content selection unit for selecting content from the selected cluster based on a similarity between one or more contents previously selected from the selected cluster and one or more contents not previously selected from the selected cluster; It is possible to provide a content selection server, including a content providing unit for providing the selected content to a user terminal, wherein the selected content corresponds to any one of a plurality of preset famous content.

Another embodiment of the present invention provides a computer program stored in a computer-readable recording medium including a sequence of instructions for selecting content, wherein the computer program is executed by a computing device to provide a plurality of clusters for a plurality of content. , define a first cluster set including the formed plurality of clusters, and select a cluster from the first cluster set based on the similarity between the plurality of clusters and the second cluster set included in the first cluster set and including the selected cluster in the second cluster set, and from the selected cluster based on the similarity between one or more contents previously selected from the selected cluster and one or more contents not previously selected from the selected cluster. A computer program stored in a computer-readable recording medium comprising a sequence of instructions for selecting content and providing the selected content to a user terminal, wherein the selected content corresponds to any one of a plurality of preset famous content can provide

The above-described problem solving means are merely exemplary, and should not be construed as limiting the present invention. In addition to the exemplary embodiments described above, there may be additional embodiments described in the drawings and detailed description.

According to any one of the above-described problem solving means of the present invention, a content cluster can be formed based on the degree of similarity between the contents, and a cluster having a low similarity to a previously selected cluster can be selected from the formed content cluster, and the selected cluster can be selected. It is possible to provide a content selection method capable of providing a content set each including different types of content by selecting content having a low similarity to previously selected content.

That is, by providing a content set composed of various contents, it is possible to provide a content selection method that can systematically identify users' tastes and preferences even with a small amount of feedback.

1 is an exemplary diagram for explaining a conventional content recommendation system according to an embodiment of the present invention.
2 is a block diagram of a content selection server according to an embodiment of the present invention.
3 is an exemplary diagram of a flow of a content selection system according to an embodiment of the present invention.
4 is an exemplary diagram of a plurality of clusters according to an embodiment of the present invention.
5 is a flowchart of a content selection method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is "connected" with another part, this includes not only the case of being "directly connected" but also the case of being "electrically connected" with another element interposed therebetween. . Also, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless otherwise stated, and one or more other features However, it is to be understood that the existence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded in advance.

In this specification, a "part" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. In addition, one unit may be implemented using two or more hardware, and two or more units may be implemented by one hardware.

Some of the operations or functions described as being performed by the terminal or device in this specification may be instead performed by a server connected to the terminal or device. Similarly, some of the operations or functions described as being performed by the server may also be performed in a terminal or device connected to the corresponding server.

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

2 is a block diagram of a content selection server according to an embodiment of the present invention. Referring to FIG. 2 , the content selection server 200 includes a content clustering unit 210 , a cluster defining unit 220 , a cluster selecting unit 230 , a cluster processing unit 240 , a content selecting unit 250 , and a content providing unit. (260). The components that can be controlled by the content selection server 200 are illustrated by way of example.

Each component of the content selection server 200 of FIG. 2 is generally connected through a network. For example, as shown in FIG. 2 , the content clustering unit 210 , the cluster defining unit 220 , the cluster selecting unit 230 , the cluster processing unit 240 , the content selecting unit 250 , and the content providing unit 260 . ) can be connected simultaneously or at intervals of time.

A network refers to a connection structure that allows information exchange between each node, such as terminals and servers, and includes a local area network (LAN), a wide area network (WAN), and the Internet (WWW: World). Wide Web), wired and wireless data communication networks, telephone networks, wired and wireless television networks, and the like. Examples of wireless data communication networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi, Bluetooth communication, infrared communication, ultrasound communication, Visible Light Communication (VLC), LiFi, and the like, but are not limited thereto.

3 is an exemplary diagram of a flow of a content selection system according to an embodiment of the present invention. Referring to FIG. 3 , the content selection server 200 may be connected to the content database DB 310 and the API server 320 , and the content and content selected to the user terminal 330 through the API server 320 . A set (d) can be provided. For example, the content selection server 200 receives content from the content database 310 , selects a content set composed of various types of content, and selects the content selected for the user terminal 330 through the API server 320 . A set can be provided.

For example, the content database 310 may store and manage the collected content. In addition, the content database 310 may store and manage the meta information of the collected content.

For example, the content selection server 200 may receive content and content meta information (a) from the content database 310 and perform content embedding, clustering, and selection. The content selection server 200 may generate the entire cluster set, and may update (b) the generated cluster set to the API server 320 every day.

For example, when receiving the terminal information c from the user terminal 330 , the API server 320 may provide the selected content and the content set d to the user terminal 330 .

For example, the user terminal 330 may transmit terminal information and information (c) about the pre-selected content set to the API server 320, and receive the content set (d) from the API server 320 . can The user terminal 330 may display the received content set d.

The content selection server 200 according to an embodiment of the present invention may provide a content set composed of various types of content to each user terminal 330 in real time through the flow of the content selection system described above. Hereinafter, a method of selecting content through each department configured in the content selection server 200 will be described in detail.

First, the content selection server 200 according to an embodiment of the present invention may collect content from various sources, and may extract a famous content list from among the collected content. For example, the popular content list may include content rated by multiple users, content viewed by multiple users, and content recognized by multiple users.

The content selection server 200 may remove duplicate content from the extracted famous content list and classify and include famous content in each of a plurality of clusters to be described later. That is, each of the plurality of clusters may include famous content.

The content clustering unit 210 according to an embodiment of the present invention may form a plurality of clusters for a plurality of content. For example, the content clustering unit 210 may embed the collected content into an i-dimensional vector to form a plurality of clusters. The content clustering unit 210 may process each collected content as a document and word-process each meta-information of the content.

The content clustering unit 210 may embed the collected content to bring content with high similarity closer to the vector space. For example, the content clustering unit 210 may embed all the collected content into an i-dimensional vector using the Doc2Vec algorithm. The Doc2Vec algorithm can generate vectors in units of sentences, paragraphs, or documents.

The content clustering unit 210 may apply a Paragraph/Document embedding method other than the Doc2Vec algorithm to the collected content. For example, the content clustering unit 210 may mathematically express the relationship between content through content embedding processing.

4 is an exemplary diagram of a plurality of clusters according to an embodiment of the present invention. Referring to FIG. 4 , the content embedded in the content clustering unit 210 may form a cluster between similar content, respectively.

For example, the content clustering unit 210 generates a plurality of clusters ((a), (b), (c), (d), (e), (f), (g)) based on the similarity between the contents. can be formed Specifically, the content clustering unit 210 may correspond the entire content to the i-dimensional vector space. Thereafter, the content clustering unit 210 may form the entire content into clusters (a) to (g) by using a spherical K-means algorithm. The content clustering unit 210 may form the entire content into a plurality of clusters by using a cosine distance-based clustering algorithm other than the spherical K-means algorithm. The content clustering unit 210 may identify a content cluster that is aggregated in the vector space through a clustering algorithm.

Referring to FIG. 4 , similarity between contents may be confirmed based on the degree of aggregation of specific contents in an i-dimensional vector space. For example, in the case of '007 series', it includes common features such as 'spy (genre)', 'action (genre)', 'James Bond (character)', 'original novel', 'M16 (material)', etc. are doing Therefore, the '007 series' may have an aggregated form in the vector space.

The cluster defining unit 220 according to an embodiment of the present invention may define a first cluster set including a plurality of formed clusters.

For example, the cluster defining unit 220 may define all of the formed clusters as the first cluster set, and define the empty set as the second cluster set. Specifically, the cluster definition unit 220 may define all of the plurality of clusters as the first cluster set, and define a cluster previously selected from the first cluster set as the second cluster set. That is, the first and second cluster sets defined by the cluster definition unit 220 are not a cluster based on similarity between contents, but a set of concepts defined for performing a content selection algorithm.

The cluster selector 230 according to an embodiment of the present invention may select a cluster from the first cluster set based on a similarity between a plurality of clusters included in the first cluster set and the second cluster set. For example, the cluster selector 230 may select a cluster having a low similarity to the previously-selected cluster from the first cluster set through Equation 1 below.

<Equation 1>

는 집합 A에서 집합 B를 제외한 차집합으로서, 기 선출되지 않은 군집의 집합을 의미한다. 군집 S는 기 선출되지 않은 군집 집합에 포함되어 있는 군집이고, S^*는 기 선출되지 않은 군집 집합으로부터 선출된 군집을 의미한다. In Equation 1, A means a set for the entire cluster, and B means a set for a pre-selected cluster (cluster b). therefore,

is the difference set excluding set B from set A, which means a set of non-pre-selected clusters. Cluster S is a cluster included in a non-pre-elected cluster set, and S ^* means a cluster selected from a non-pre-elected cluster set.

That is, in Equation 1, A may mean a first cluster set, B may mean a second cluster set, and S ^* is a pre-selected cluster set (second cluster set) and a non-pre-elected cluster set. It may mean a cluster selected from a non-pre-selected cluster set (the first cluster set) based on the similarity between the sets (the first cluster set).

는 군집 S의 11p중심 벡터와 군집 b의 중심 벡터 간의 코사인 유사도를 의미한다. argmin은 군집 S(기 선출되지 않은 군집)와 군집 b(기 선출된 군집)의 유사도에 기초하여 유사도 합이 가장 적은 군집을 선출함을 의미한다. In addition, in Equation 1, centroid _s means the center vector of the cluster S,

denotes the cosine similarity between the 11p centroid vector of cluster S and the centroid vector of cluster b. argmin means that the cluster with the smallest sum of similarities is selected based on the similarity between cluster S (un-elected cluster) and cluster b (pre-elected cluster).

That is, the cluster selector 230 may calculate the sum of similarities between the non-pre-elected cluster and the pre-selected cluster through Equation (1).

The cluster selector 230 may select a cluster having a low similarity to the second cluster set B from the first cluster set A through Equation 1 . That is, the cluster selector 230 may select a cluster having a low similarity to the previously selected cluster set from the non-pre-selected cluster set through Equation 1 .

When the size of the second cluster set B is 0, the cluster selector 230 according to an embodiment of the present invention may select any one cluster from the first cluster set A. For example, when there is no pre-selected cluster, the cluster selector 230 may randomly select a cluster from the first cluster set (A).

For another example, when there is a pre-selected cluster, that is, when the size of the second cluster set B is not 0, the cluster selector 230 determines the first cluster set A based on Equation 1 It is possible to select a cluster with a low similarity to the previously elected cluster from

The cluster selector 230 selects a cluster having a low similarity to a previously selected cluster from the first cluster set based on the similarity between the plurality of clusters included in the first cluster set (A) and the second cluster set (B). Thus, it is possible to distinguish a cluster that is a target of content selection from a previously selected cluster.

The cluster processing unit 240 according to an embodiment of the present invention may include the selected cluster in the second cluster set. For example, the cluster processing unit 240 includes the cluster (S ^* ) selected from the first cluster set (A) by the cluster selector 230 as the second cluster set (B) to include the pre-selected cluster set (B). ) can be updated.

The content selection server 100 according to an embodiment of the present invention may further include a cluster redefining unit (not shown). When the number of clusters included in the first cluster set (A) is the same as the number of clusters included in the second cluster set (B), the cluster redefiner may redefine the second cluster set (B) as an empty set.

For example, when the number of clusters included in the first cluster set (A) is equal to the number of clusters included in the second cluster set (B), that is, when the cluster redefining unit traverses the entire cluster set once , it is possible to redefine the second cluster set (B) as an empty set. As another example, when the number of clusters included in the first cluster set (A) is equal to the number of clusters included in the second cluster set (B), the cluster redefining unit may increase the number of iterations by 1.

The content selection unit 250 according to an embodiment of the present invention includes one or more contents (d) previously selected from the selected cluster (S ^* ) and one or more contents (c) not previously selected from the selected cluster (S ^* ). ), content can be selected from the selected cluster (S ^* ) based on the similarity between them.

The selected content (d) according to an embodiment of the present invention may correspond to any one of a plurality of preset famous content. For example, the content selection unit 250 may select content having a low similarity to one or more previously selected content from the selected cluster through Equation 2 below.

<Equation 2>

는 선출된 군집에서 유명 콘텐츠만으로 구성된 집합에서 기 선출된 유명 콘텐츠를 제외한 차집합으로 기 선출되지 않은 유명 콘텐츠를 의미한다.

는 선출된 군집에서 유명 콘텐츠만으로 구성된 집합에서 기 선출된 유명 콘텐츠를 의미한다. In Equation 2, S ^* _G means a set consisting of only famous content from the elected cluster (S ^* ),

? means famous content that is not pre-elected from the set consisting of only famous content in the elected cluster except for the well-known content previously elected in the set.

denotes famous content previously selected from a set consisting of only famous content from the selected cluster.

는 기 선출되지 않은 콘텐츠(c)와 기 선출된 콘텐츠(d) 간의 코사인 유사도를 의미하고, c^*는 기 선출되지 않은 콘텐츠와 기 선출된 콘텐츠 간의 유사도에 기초하여 기 선출되지 않은 콘텐츠 중에서 선출된 콘텐츠를 의미한다.That is, in Equation 2, c denotes content that has not been previously selected, and d denotes content that has been previously selected.

denotes the cosine similarity between the non-pre-selected content (c) and the pre-selected content (d), and c ^* denotes the cosine similarity between the non-pre-selected content and the pre-selected content based on the similarity between the non-pre-elected content and the pre-selected content. means content.

For example, the content selector 250 may select the content based on the similarity between the one or more contents previously selected from the group S ^* selected through Equation 2 and the one or more contents not previously selected.

The content selection unit 250 selects the content based on the similarity between the one or more content (d) previously selected from the selected cluster (S ^* ) and the one or more content (c) not previously selected by selecting the content and Distinctive content can be selected. That is, the content selection unit 250 may generate a content set composed of various types of content by selecting the content c having a low similarity to the previously selected content d.

On the other hand, the content selection unit 250 according to an embodiment of the present invention, when there is no content (d) previously selected from the selected cluster (S ^* ), any content corresponding to famous content from the selected cluster (S ^* ) One content can be selected. For example, when there is no pre-selected content d, the content selector 250 may randomly select any one of famous content from the selected cluster S ^* .

인 경우, 즉, 선출된 군집 S^*에 유명 콘텐츠가 더 이상 존재하지 않는 경우, 선출된 군집 S^*를 기 선출되지 않은 군집 집합, 제 1 군집 집합(A)에서 제외시켜, 제 1 군집 집합(A)을 갱신할 수 있다. In addition, when all contents corresponding to famous contents are selected from the selected cluster (S ^* ) according to an embodiment of the present invention, the selected cluster may be excluded from the first cluster set (A). For example, the content selection unit 250 may

In the case of , that is, when famous content no longer exists in the elected cluster S ^* , the elected cluster S ^* is excluded from the previously unassigned cluster set, the first cluster set (A), and the first cluster set ( A) can be updated.

According to an embodiment of the present invention, a cluster is selected from the first cluster set (A) until the number of selected content (c*) becomes equal to the number of preset content, and the selected cluster is set to a second cluster set ( B), and the content selection from the elected cluster (S ^* ) can be repeated.

That is, until all the contents included in the entire cluster set are selected, based on Equations 1 and 2 described above, a cluster having a low similarity to the previously elected cluster is selected from the entire cluster set, and previously selected from the selected cluster. It is possible to select content with a low similarity to the content that has been used.

The content providing unit 260 according to an embodiment of the present invention may provide the selected content to the user terminal. For example, the content providing unit 260 may provide a content set composed of content having a low similarity to the user terminal. That is, the content providing unit 260 may provide a content set composed of various content to the user terminal.

5 is a flowchart of a content selection method according to an embodiment of the present invention. The content selection method shown in FIG. 5 includes steps processed in time series according to the embodiments shown in FIGS. 1 to 4 . Therefore, even if omitted below, it is also applied to the method of selecting content in the content selection server according to the embodiment shown in FIGS. 1 to 4 .

In step S510, the content selection server may form a plurality of clusters for a plurality of content.

In step S520, the content selection server may define a first cluster set including the formed plurality of clusters.

In step S530, the content selection server may select a cluster from the first cluster set based on the similarity between the plurality of clusters included in the first cluster set and the second cluster set.

In step S540, the content selection server may include the selected cluster in the second cluster set.

In step S550 , the content selection server may select content from the selected cluster based on the similarity between one or more contents previously selected from the selected cluster and one or more contents not previously selected from the selected cluster.

In step S560, the content selection server may provide the selected content to the user terminal.

In the above description, steps S510 to S560 may be further divided into additional steps or combined into fewer steps according to an embodiment of the present invention. In addition, some steps may be omitted as necessary, and the order between the steps may be switched.

The method for selecting content in the content selection server described through FIGS. 1 to 5 is also implemented in the form of a computer program stored in a computer-readable recording medium executed by a computer or a recording medium including instructions executable by the computer. can be Also, the method for selecting content in the content selection server described with reference to FIGS. 1 to 5 may be implemented in the form of a computer program stored in a computer-readable recording medium that is executed by a computer.

A computer-readable recording medium may be any available medium that can be accessed by a computer, and includes both volatile and nonvolatile media, removable and non-removable media. In addition, the computer-readable recording medium may include a computer storage medium. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

The above description of the present invention is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and likewise components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be interpreted as being included in the scope of the present invention. do.

200: content selection server
210: content cluster
220: Cluster Justice
230: cluster electoral division
240: swarm processing unit
250: content selection unit
260: content provider

Claims (13)

A content selection method performed by a content selection server, the method comprising:
forming a plurality of clusters for a plurality of contents;
defining a first cluster set including the formed plurality of clusters;
selecting a cluster from the first cluster set based on a degree of similarity between a plurality of clusters and a second cluster set included in the first cluster set;
including the selected cluster in the second cluster set;
selecting content from the selected cluster based on a similarity between one or more contents previously selected from the selected cluster and one or more contents not previously selected from the selected cluster;
providing the selected content to a user terminal
including,
The content selection method, wherein the selected content corresponds to any one of a plurality of preset famous content.
The method of claim 1,
The step of selecting a cluster from the first cluster set includes:
When the size of the second cluster set is 0, any one cluster is selected from the first cluster set.
The method of claim 1,
The step of selecting content from the selected cluster comprises:
When there is no pre-selected content from the selected cluster, any one content corresponding to the famous content is selected from the selected cluster.
4. The method of claim 3,
excluding the selected cluster from the first cluster set when all the contents corresponding to the famous content are selected from the selected cluster
Which will further include, the content selection method.
The method of claim 1,
redefining the second cluster set as an empty set when the number of clusters included in the first cluster set is equal to the number of clusters included in the second cluster set
Further comprising, a content selection method.
The method of claim 1,
selecting the cluster from the first cluster set until the number of the selected content is equal to the number of the preset content, including the selected cluster in the second cluster set, and from the selected cluster The content selection method is to repeat the step of selecting the content.
In the content selection server,
a content clustering unit for forming a plurality of clusters for a plurality of content;
a cluster defining unit defining a first cluster set including the formed plurality of clusters;
a cluster selector configured to select a cluster from the first cluster set based on a degree of similarity between a plurality of clusters and a second cluster set included in the first cluster set;
a cluster processing unit for including the selected cluster in the second cluster set;
a content selection unit for selecting content from the selected cluster based on a similarity between one or more contents previously selected from the selected cluster and one or more contents not previously selected from the selected cluster;
A content providing unit that provides the selected content to a user terminal
including,
The content selection server, wherein the selected content corresponds to any one of a plurality of preset famous content.
8. The method of claim 7,
The group selection unit,
When the size of the second cluster set is 0, the content selection server for selecting any one cluster from the first cluster set.
8. The method of claim 7,
The content selection unit,
If there is no pre-selected content from the selected cluster, the content selection server to select any one content corresponding to the famous content from the selected cluster.
10. The method of claim 9,
When all the contents corresponding to the famous content are selected from the selected cluster, the selected cluster is excluded from the first cluster set.
8. The method of claim 7,
When the number of clusters included in the first cluster set is the same as the number of clusters included in the second cluster set, the cluster redefining unit redefines the second cluster set as an empty set.
Further comprising, the content selection server.
8. The method of claim 7,
The cluster is selected from the first cluster set until the number of the selected content is equal to the number of the preset content, the selected cluster is included in the second cluster set, and the content from the selected cluster A content election server, which repeats the election.
A computer program stored on a computer-readable recording medium comprising a sequence of instructions for selecting content, the computer program comprising:
When the computer program is executed by a computing device,
forming a plurality of clusters for a plurality of contents,
defining a first cluster set including the formed plurality of clusters;
selecting a cluster from the first cluster set based on a degree of similarity between a plurality of clusters and a second cluster set included in the first cluster set;
including the elected cluster in the second cluster set,
selecting content from the selected cluster based on a similarity between one or more contents previously selected from the selected cluster and one or more contents not previously selected from the selected cluster;
Provide the selected content to the user terminal,
wherein the selected content includes a sequence of instructions to correspond to any one of a plurality of preset popular content.

Images