KR102181835B1 - Method for determining node with high influence, apparatus for the same, method for minimizing marketing cost in viral marketing and system for the same - Google Patents

Abstract

A method for determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing the same, wherein the method for determining a high-impact node includes at least one neighboring node corresponding to a target node. Determining, determining at least one total contributing node corresponding to each of the at least one neighboring node, and determining a neighboring node having the highest influence from the at least one neighboring node based on the total contributing node. Can include.

Description

How to determine the high-impact node, the device for determining the high-impact node, the method for minimizing marketing costs in viral marketing, and a system for performing it (METHOD FOR DETERMINING NODE WITH HIGH INFLUENCE, APPARATUS FOR THE SAME, METHOD FOR MINIMIZING MARKETING COST IN VIRAL MARKETING AND SYSTEM FOR THE SAME}

A method for determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing the same.

Viral marketing refers to a marketing method in which publicity/unpromotional information about an event, product, or company is spread among people through rumors among people. Such viral marketing can be performed through a medium capable of propagating, and in recent years, various types/types of social networks due to the development of information and communication technologies are used as a major platform for such viral marketing. Such social networks include, for example, a home page, a blog, a micro blog (such as Facebook, Twitter, or Instagram), an internet cafe, a content hosting platform (for example, an internet broadcasting platform such as YouTube, etc.), and /Or Internet messenger etc. exist.

A very large number of users participate in the social network, and among them, there are users (hereinafter referred to as seed users) that have a relatively stronger and more influence on other users. For example, celebrities such as celebrities have relatively more neighbors, followers, and subscribers than other common users, and their created or uploaded posts (symbols, text, shapes, still images and/or Or, it may include a video) is transmitted or shared to a relatively larger number of other users.

In other words, the activation of the seed user's node (such as the seed user's account or posts) can activate a large number of other users (or other user's nodes). Here, the activation of the seed user's node means performing an action, such as posting a post, and the activation of another user's node means that all or part of the seed user's posts by another user. It means viewing/copying/quoting/sharing/delivery.

The conventional method of maximizing profits for viral marketing has been to find a target user who affects a large number of users of the maximum or a certain level or higher. In other words, in order to maximize profits, income must be maximized or cost must be minimized, but the conventional method was to inform as many users as possible and search for target users who can maximize income. This method was performed by measuring the maximum number of users that can be activated by the active target user. However, this method has a problem on the premise that the target user is active.

A method for determining a high-impact node that can determine at least one neighboring node with high influence over a target node within an appropriate range, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing this What is provided is a task to be solved.

In order to solve the above problems, a method for determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing the same are provided.

The method of determining a high impact node includes determining at least one neighboring node corresponding to a target node, determining at least one total contributing node corresponding to each of the at least one neighboring node, and the total contributing node. As a result, it may include determining a node having the highest influence from the at least one neighboring node.

The determining of at least one total contributing node corresponding to each of the at least one neighboring node may include determining at least one first confirmation target node corresponding to the neighboring node, and the at least one first confirmation target node Selecting at least one first currently active node affecting the at least one neighboring node from, adding the at least one first currently active node to at least one contributing node corresponding to the at least one neighboring node Thus, it may include updating the contributing node corresponding to the neighboring node.

Selecting the at least one first active node affecting the at least one neighboring node from the at least one first verification target node may include the at least one first verification target node based on a predefined probability It may include the step of selecting at least one of the first currently active node.

The determining of at least one total contributing node corresponding to each of the at least one neighboring node may include searching for a first new active node among the at least one first currently active node, wherein the first new active node is the contributing node. It may further include the step of including a node that does not overlap with the node.

The determining of the at least one total contributing node corresponding to each of the at least one neighboring node may include, if the first new active node is not found among the at least one first currently active node, the contributing node is the total contribution. It may further include determining a node.

The determining of at least one total contributing node corresponding to each of the at least one neighboring node may include determining at least one second verification target node corresponding to each of the searched first new active nodes, the at least one Selecting at least one second currently active node affecting the first new active node from a second verification target node, adding the at least one second currently active node to the at least one contributing node to the neighbor The method may further include updating a contributing node corresponding to the node, and searching for a second new active node from among the at least one second currently active node.

The method of determining the high impact node may further include initializing the contributing node, wherein the contributing node may be set and initialized in the same manner as the first new active node.

Determining the neighboring node having the highest influence from the at least one neighboring node based on the total contributing node includes obtaining each of the influences of the at least one neighboring node on the target node, wherein the at least The influence of one neighboring node may be determined as an inverse number of total contributing nodes corresponding to the neighboring node.

Determining a group of neighboring nodes based on the total contributing nodes, wherein the group of neighboring nodes includes a minimum number of neighboring nodes having the most influence, and at least one normalized influence is obtained by normalizing the at least one influence. It may include the step of obtaining.

In the step of normalizing the at least one influence to obtain at least one normalized influence, calculating a ratio between the difference between the influence of the neighboring node and the minimum influence on the target node and the difference between the maximum influence and the minimum influence By doing so, it may include the step of normalizing the influence of the neighboring node on the target node.

The determining of the neighboring node having the highest influence from the at least one neighboring node based on the total contributing node may include a neighboring node having a relatively high influence among the at least one neighboring node based on the at least one total contributing node. And selecting a high-impact neighbor node having a relatively high influence among at least one neighboring node other than the high-impact node, based on the at least one total contributing node.

The determining of the neighboring node having the highest influence from the at least one neighboring node based on the total contributing node may include calculating at least one influence for each of the at least one neighboring node corresponding to the target node, Obtaining at least one normalized influence with respect to the at least one influence, determining a normalized influence having the largest value among the at least one normalized influence, and detecting a neighbor node corresponding to the determined normalized influence Determining a neighboring node having the highest influence, wherein the value of the influence may be calculated by obtaining an inverse of the contributing node.

Determining a neighboring node having the highest influence from the at least one neighboring node based on the total contributing node is when the sum of the at least one normalized influence is equal to or greater than a predefined threshold value It may further include the step of detecting the neighboring node having the highest influence to.

The method of determining a high influence node may further include obtaining a group of neighboring nodes including the neighboring nodes having the highest influence.

The set of neighboring nodes may include neighboring nodes having the highest influence less than or equal to a predefined threshold.

The apparatus for determining a high-influence node, after determining an information collecting unit for collecting information related to at least one neighboring node corresponding to a target node and at least one total contributing node corresponding to each of the at least one neighboring node, It may include a processor that determines a neighboring node having the highest influence from the at least one neighboring node based on the contributing node.

The processor determines at least one verification target node corresponding to the neighboring node, selects at least one currently active node affecting the at least one neighboring node from the at least one verification target node, and the at least At least one total contributing node corresponding to each of the at least one neighboring node by updating the contributing node corresponding to the neighboring node by adding the at least one currently active node to at least one contributing node corresponding to one neighboring node Can be determined.

The processor, based on the at least one total contributing node, selects a neighboring node having a relatively high influence among the at least one neighboring node, and based on the at least one total contributing node, at least another node other than the high influence node A highly influential neighboring node having a relatively high influence may be selected from one neighboring node, and a neighboring node having the highest influence may be determined from the at least one neighboring node based on the total contributing node.

The marketing cost minimization determination method may be performed based on a social network including at least one target user and at least one neighboring user corresponding to at least one target user, and at least one neighbor corresponding to at least one target user Determining a user, determining at least one total contributing user corresponding to each of the at least one neighboring user, and having the highest influence on the target user from the at least one neighboring user based on the total contributing user It may include determining a neighbor user.

The marketing cost minimization system includes a social network device having information on at least one target user and information on at least one neighboring user corresponding to the target user, and communicatively connected with the social network device, and the at least one And an apparatus for determining at least one total contributing user corresponding to each of the neighboring users, and determining a neighboring user having the highest influence from the at least one neighboring user based on the total contributing user.

According to the above-described method for determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing the same, at least one neighboring node having a high influence on the target node is within the required limit You can get the effect of being able to make an appropriate decision.

According to the above-described method of determining a high-influence node, a device for determining a high-influence node, a method for minimizing marketing costs in viral marketing, and a system for performing the same, a high-influence neighbor node is detected, without the detection of a large number of neighbor nodes. The effect of activating the target node to a desired level can also be obtained.

According to the above-described method of determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing this, the effect of cost reduction by optimally selecting a neighboring node required for activation of the target node. You can also get.

According to the above-described method of determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing this, the profits that can be obtained through viral marketing are minimized by minimizing the cost of viral marketing You can get the effect that can be maximized.

According to the above-described method of determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and a system for performing this, the user who has the most influence on the spread of content in the social network is selected and is The effect of being able to promote a product or service within the metropolitan network can also be obtained.

Using the above-described high-impact node determination method, high-impact node determination device, a method of minimizing marketing costs in viral marketing, and a system for performing this, online promotion/advertising costs are determined through social networks, and cost-effective advertising agencies It is also possible to obtain the effect of more efficiently and appropriately performing the decision of the product and/or the prediction of the profits that can be obtained from the social network before product launch, and also more effectively and accurately perform the cost-benefit-analysis. You can also get the effect of being able to do it.

1 is a diagram showing an embodiment of an apparatus for determining a high impact node.
2 is a diagram illustrating an embodiment of a target node and neighboring nodes connected from the target node.
3 is a diagram illustrating an embodiment of an activated neighbor node.
4 is a diagram illustrating an embodiment of a node to be checked for activated neighboring nodes.
5 is a diagram showing an embodiment of a node to be checked for a current active node.
6 is a diagram illustrating an embodiment of a method of determining a neighboring node having the highest influence from at least one neighboring node.
7 is a first diagram illustrating an embodiment of a method of determining a neighboring node having the highest influence from at least one neighboring node.
8 is a second diagram illustrating an embodiment of a method of determining a neighboring node having the highest influence from at least one neighboring node.
9 is a third diagram illustrating an embodiment of a method of determining a neighboring node having the highest influence from at least one neighboring node.
10 is a diagram showing an embodiment of an algorithm.
11 is a diagram illustrating an embodiment of a system for minimizing marketing costs in viral marketing.
12 is a diagram illustrating an embodiment of a network in which viral marketing is performed.
13 is a first flowchart for explaining an embodiment of a method for determining a high impact node.
14 is a second flowchart illustrating an embodiment of a method for determining a high impact node.
15 is a third flowchart for explaining an embodiment of a method for determining a high impact node
16 is a flowchart illustrating an embodiment of a method for determining the minimization of marketing costs.

In the following specification, the same reference numerals refer to the same elements unless otherwise specified. The term "unit" used below may be implemented as software or hardware, and according to an embodiment, the term "unit" is implemented as one part, or one "unit" is implemented as a plurality of parts. It is also possible.

When a part is said to be connected to another part throughout the specification, it may mean a physical connection depending on the part and another part, or may mean electrically connected. In addition, when a part includes another part, this does not exclude another part other than the other part unless otherwise stated, and it means that another part may be included further according to the designer's choice. do.

Terms such as first and second are used to distinguish one part from another, and unless otherwise specified, they do not mean sequential expressions. Singular expressions may include plural expressions, unless the context clearly has exceptions.

Hereinafter, an embodiment of an apparatus for determining a high impact node will be described with reference to FIGS. 1 to 10.

1 is a diagram showing an embodiment of an apparatus for determining a high impact node, and FIG. 2 is a diagram showing an embodiment of a network consisting of a plurality of nodes.

1 and 2, the apparatus 100 for determining a high-impact node is communicatively connected with the analysis target network 10 to obtain information from the analysis target network 10 and obtain the obtained information. Based on the relatively high influence on the target node (v) among the nodes (u1, u2, u3, u4, hereinafter neighboring nodes) having a relationship with a predetermined node (hereinafter referred to as the target node (v)) of the analysis target network 10 It may be designed to determine at least one node (hereinafter, a high impact node) having.

The apparatus 100 for determining a high impact node can be implemented using a computing device capable of processing information. A computing device capable of processing information may be implemented using, for example, a desktop computer, a laptop computer, or a server dedicated computer, and according to an embodiment, a portable device such as a smart phone, a tablet PC, or a portable game machine, etc. , Internet of Things-based digital televisions or electronic devices such as refrigerators may be used. Further, in addition to these, various electronic devices capable of calculating and processing may be used as an example of the apparatus 100 for determining a high-influence node.

Referring to FIG. 2, the analysis target network 10 may include a plurality of nodes (v, u, a, b and/or c, etc.). Here, a predetermined relationship (link) may be formed between at least two nodes among the plurality of nodes (v, u, a, b and/or c, etc.). For example, at least one node (v, u1 to u5, a1 to a12, b1 to b5, c1 to a plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) At least one of c3) may be connected to at least one of other nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) by a predetermined relationship. In this case, any one node (for example, the fourth neighboring node u4) among the plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) is only one node. (For example, it is possible to establish a relationship with a4), and/or any one node (for example, the first neighboring node u1) is one node is a plurality of nodes (for example, v, a8 to a10, etc.).

The relationship between at least two nodes among a plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) may include a physical relationship between each other or a virtual relationship. have. Virtual relationships may include relationships such as friends, neighbors, following or subscriptions on a social network, for example.

When a relationship exists between a plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.), a plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) may exert an influence (influence) in one or both directions. In this case, each node (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) is assigned to another node (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) The intensity of the influence on each other may be different. In other words, a particular node (e.g., target node (v)) may be strongly or weakly affected by other nodes (e.g., first neighboring node (u1)) that have a relationship with it. Yes, or it may not be affected at all. In addition, such influence may be variable depending on changes in platform, surrounding environment, situation and/or time.

The operation of any one node (eg, the first neighboring node u1) may induce the operation of another node (eg, the target node v). Whether or not the other node v operates due to the operation of one node u1 may depend on the intensity of the above-described influence. For example, a certain node (for example, a first neighbor node (u1)) is connected to a specific node (for example, a target node (v)) and has a relatively strong influence on a specific node (v). In this case, when a predetermined node u1 is activated, a specific node v may be activated with a high probability in response thereto. Of course, even if the node u1 having a relatively strong influence is activated, the specific node v may not be activated in probability. Conversely, when a node (e.g., a second neighboring node (u2)) that has a relatively weak influence on a specific node (for example, the target node (v)) is activated, the specific node (v) is activated with a low probability. Can be.

으로 표현되고, 특정한 노드(v)와 무관한 노드들(예를 들어, a6)의 집합을

으로 표현될 수 있다. 이 경우,

및

은 각각

및

으로 정의될 수 있다.The analysis target network 10 may be mathematically expressed as G(V, E). Here, G(.) means a network, V means a set of node(s) (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) in the network to be analyzed 10, E denotes relationships within the network 10 to be analyzed. In addition, a set of node(s) (u1 to u4, etc.) having a relationship with each other for a specific node (for example, target node (v))

And a set of nodes (e.g., a6) not related to a specific node (v)

It can be expressed as in this case,

And

Is each

And

Can be defined as

According to an embodiment, the apparatus 100 for determining a high impact node includes an information collection unit 110 that collects and obtains necessary information from the node information providing apparatus 190, and the information collected by the information collection unit 110. It may include a processor 120 that collects from the information collection unit 110 and processes and controls the overall operation of the apparatus 100 for determining an influence node.

As shown in FIG. 2, the information collection unit 110 receives information from a network 10 including a plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.). Can be collected. Specifically, the information collection unit 110 may obtain information on a plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) in the network 10. In addition, the information collection unit 110 obtains information on a relationship formed between at least two of the plurality of nodes (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) of the network 10 You may.

The information collection unit 110 is, for example, a wired/wireless communication module capable of communicating with the analysis target network 10 or an external storage medium having information on the analysis target network 10 (for example, It can be implemented using a physical interface (e.g., an ATA (Advanced Technology Attachment) interface, a universal serial bus interface and/or a Thunderbolt interface, etc.) that can be connected to a semiconductor memory device or a magnetic disk storage device. In addition, a semiconductor chip (eg, a communication chip) for processing information received as necessary may be further included. Here, the wired/wireless communication module means a communication module that can access a wired communication network, a wireless communication network, or a combination thereof, and the wireless communication network includes at least one of a short-range communication and a long-distance communication network (mobile communication network). I can.

The processor 120 uses nodes on the network 10 (v, u1 to u5, a1 to a12, b1 to b5, c1 to c3, etc.) and a relationship therebetween, and a predetermined node (hereinafter, a target node ( v)), from among nodes (hereinafter referred to as neighboring nodes u1, u2, u3, u4, u5, etc.), at least one node having a relatively high influence (hereinafter, a high influence node) may be determined.

Specifically, when at least one neighboring node (u1 to u5) is connected to at least one target node (v) as shown in FIG. 2, the processor 120 is at least one of the at least one neighboring node (u1 to u5). It is provided to determine a total contributing node corresponding to one, and to determine at least one high-influence node that has the highest influence on at least one target node (v) based on the determined total contributing node. Accordingly, the processor 120 can obtain a set of neighboring nodes (hereinafter, a set of high-affect nodes) composed of high-affect nodes.

According to an embodiment, the processor 120 may determine a high impact node and determine a high impact node set using a heuristic mixed model (HM Model). In this case, the heuristic mixed model is a reverse IC model modified from an independent cascade model (IC model) and an inverse linear threshold modified from a linear threshold model (LT model). It may be implemented by employing a combination of a model (Reverse LT model) and a greedy optimization technique. Usually, the independent cascade model is used to find out how many nodes can be activated by a particular node. On the other hand, in the inverse independent cascade model, it is possible to determine how many activated nodes are required for activation of a specific node (ie, target node v) by using the independent cascade model in reverse order.

According to an embodiment, the processor 120 may include a total contributing node determining unit 122 and a high-impact node obtaining unit 124, as shown in FIG. 1. The total contribution node determination unit 122 and the high-impact node acquisition unit 124 may be logically separated and/or physically separated. When physically separated, each of the total contributing node determination unit 122 and the high-impact node acquisition unit 124 may be implemented by different semiconductor chips.

The total contributing node determination unit 122 determines at least one total contributing node corresponding to at least one neighboring node u1 to u5 for each of the at least one neighboring node u1 to u5 corresponding to the target node v. You can decide. Depending on the embodiment, the total contributing node determination unit 122 may determine the total contributing nodes for all neighboring nodes u1 to u5 of the target node v, or some neighboring nodes of the target node v ( It is also possible to determine the total contributing nodes for u1 to u5).

The total contributing node determining unit 122 may be implemented to obtain at least one total contributing node using an inverse independent cascade model.

)라고 지칭하도록 하고, 특정한 노드(u)의 활성화 여부를 위해 확인 대상이 되는 모든 노드(들) 또는 이들의 집합을 특정 노드(u)에 대한 확인 대상 노드(

)라고 지칭하도록 한다. 예를 들어, 제1 이웃 노드(u1)에 대한 확인 대상 노드(

)는, 제1 이웃 노드(u1)와 연결된 적어도 하나의 노드(a8, a9, a10 등) 또는 이들의 집합일 수 있다. 또한, 특정한 하나의 처리 과정 내에서 활성화되고 있거나 또는 활성화되어 있는 노드(들) 또는 이들의 집합을 현 활성 노드(

)라고 지칭한다. 아울러, 기 수행된 처리 과정에서 또는 현재의 처리 과정에서, 특정한 노드(u)의 활성화에 기여하는 것으로 판단된 모든 노드(들) 또는 이들의 집합을 총 기여 노드(

)라고 지칭하도록 한다.For the convenience of description below, the newly activated node(s) or a set thereof in one specific processing process (for example, p1 to p3 in FIGS. 3 to 5) is a new active node (

), and to determine whether a specific node (u) is activated or not, all the node(s) to be checked, or a set of them, are identified for a specific node (u).

). For example, the node to be checked for the first neighboring node u1 (

) May be at least one node (a8, a9, a10, etc.) connected to the first neighboring node u1 or a set thereof. In addition, the node(s) that are being activated or activated within a particular processing process, or a set of them, are currently active nodes (

). In addition, in the previously performed process or in the current process, all node(s) or a set of nodes determined to contribute to the activation of a specific node (u) are selected as the total contributing node (

).

3 is a diagram illustrating an embodiment of an activated neighbor node.

:

내지

)를 결정할 수 있다. 적어도 하나의 이웃 노드(u: u1 내지 u5) 각각에 대한 총 기여 노드(

:

내지

)의 결정은 서로 독립적으로 또는 의존적으로 수행될 수 있다. The total contributing node determination unit 122 detects at least one neighboring node (u: u1 to u5) corresponding to the target node v, and the total contributing node for each of at least one neighboring node (u: u1 to u5) (

:

To

) Can be determined. Total contributing nodes for each of at least one neighboring node (u: u1 to u5) (

:

To

The determination of) can be performed independently or dependently of each other.

:

,

및

)만을 결정할 수도 있다(P1, 이하 제1 처리 과정). According to an embodiment, the total contributing node determination unit 122 detects the activated neighbor nodes u1, u3, and u5, as shown in FIG. 3, and each of the activated neighbor nodes u1, u3, and u5 Total Contribution Node for (

:

,

And

) May be determined (P1, hereinafter, the first processing process).

:

,

및

)의 결정을 위해서, 총 기여 노드 결정부(122)는, 먼저 하나 또는 둘 이상의 활성화된 이웃 노드(u: u1, u3, u5)에 대한 신규 활성 노드(

) 및 총 기여 노드(

)를 초기화할 수 있다. 이 때, 신규 활성 노드(

) 및 총 기여 노드(

)는, 각각 분석 대상이 되는 이웃 노드(일례로 활성화된 활성화된 이웃 노드(u: u1, u3, u5))를 이용하여 초기화될 수 있다. 예를 들어, 만약 제3 이웃 노드(u3)에 대한 총 기여 노드(

)를 결정하는 경우, 제3 이웃 노드(u3)에 대한 신규 활성 노드(

) 및 총 기여 노드(

)는 {u3}으로 결정될 수 있다. 이는 하기의 수학식 1 및 수학식 2로 표현 가능하다.Total Contributing Node(

:

,

And

) To determine, the total contributing node determination unit 122 first, a new active node for one or more activated neighbor nodes (u: u1, u3, u5) (

) And the total contributing node (

) Can be initialized. At this time, the new active node (

) And the total contributing node (

) May be initialized by using neighboring nodes to be analyzed (for example, activated neighboring nodes u: u1, u3, u5). For example, if the total contributing node to the third neighboring node (u3) (

In the case of determining ), a new active node for the third neighboring node u3 (

) And the total contributing node (

) Can be determined as {u3}. This can be expressed by Equation 1 and Equation 2 below.

[Equation 1]

[Equation 2]

4 is a diagram illustrating an embodiment of a node to be checked for activated neighboring nodes.

)를 기반으로 총 기여 노드(

)가 결정될 수 있다(p2, 이하 제2 처리 과정). 이를 위해서 현 활성 노드(

)가 획득될 수 있으며, 또한 신규 활성 노드(

)도 더 획득될 수 있다.As shown in Fig. 4 in sequence, the node to be checked (

) Based on the total contributing node (

) May be determined (p2, hereinafter, the second processing process). To do this, the currently active node (

) Can be acquired, and also a new active node (

) Can also be obtained.

)이 결정될 수 있다. 특정한 노드(u)에 대한 영향력 판단을 위해, 특정한 노드(u)에 대해 관계를 가지고 있는 모든 노드(들)이 확인 및 검토될 필요가 있다. 그러므로, 확인 대상 노드(

)는 이와 같이 특정한 노드(u)와 연결된 모든 노드(들)을 포함할 수 있으며, 모든 노드(들)의 집합일 수 있다. 이에 따라, 각 이웃 노드(u)의 활성화와 관련하여 확인 및 검토하게 되는 확인 대상 노드(

)는 하기의 수학식 3으로 표현될 수 있다. 여기서, 각 이웃 노드(u)는 신규 활성 노드(

)에 속하는 것일 수 있다.For this, first, the node(s) to be checked for each of the neighboring nodes (u1 to u5) (

) Can be determined. In order to determine the influence on a specific node u, all node(s) having a relationship to the specific node u need to be identified and reviewed. Therefore, the node to be checked (

) May include all node(s) connected to a specific node (u), and may be a set of all node(s). Accordingly, the node to be verified to be checked and reviewed in relation to the activation of each neighboring node (u) (

) Can be expressed by Equation 3 below. Here, each neighboring node u is a new active node (

) May belong to.

[Equation 3]

)는, 도 4에 도시된 바와 같이, 제3 이웃 노드(u3)에 연결된 노드(들)(a1 내지 a3)를 모두 포함하여 결정될 수 있다. 이를 수학적으로 표현하면, 확인 대상 노드(

)는 {a1, a2, a3}으로 주어질 수 있다.For example, the node to be checked for the third neighboring node u3 (

) May be determined by including all of the node(s) a1 to a3 connected to the third neighboring node u3, as shown in FIG. 4. To express this mathematically, the node to be checked (

) Can be given as {a1, a2, a3}.

)가 결정되면, 확인 대상 노드(

) 중에서 선택된 이웃 노드(u1, u3, u5)에 대한 현 활성 노드(

)가 결정될 수 있으며, 현 활성 노드(

)를 이용하여 신규 활성 노드(

) 및 총 기여 노드(

)가 갱신될 수 있다.Node to be checked (

) Is determined, the node to be checked (

), the currently active node for the selected neighboring node (u1, u3, u5) (

) Can be determined, and the currently active node (

) Using the new active node (

) And the total contributing node (

) Can be updated.

)로 분류될 수 있다. 다시 말해서, 모든 확인 대상 노드(a1 내지 a3) 중에서 소정의 이웃 노드(u3)에 영향을 미치는 것으로 판단된 확인 대상 노드(a2 및 a3)가 현 활성 노드(

)로 결정된다. 반대로 영향을 미치지 못하는 것으로 판단된 다른 확인 대상 노드(a)는 현 활성 노드(

)로 인정되지 않는다. 이는 수학적으로는

={a2 및 a3}으로 표현될 수 있다.If it is determined that at least one node (e.g., a2 and a3) among the nodes to be checked (a1 to a3) affects the activation of the connected node (for example, the third neighboring node u3), At least one node (a2 and a3) determined to be the currently active node (

) Can be classified. In other words, among all the check target nodes a1 to a3, the check target nodes a2 and a3 determined to affect a predetermined neighbor node u3 are the currently active nodes (

). Conversely, the other node to be checked (a) determined to have no effect is the currently active node (

) Is not recognized. Mathematically

It can be expressed as ={a2 and a3}.

According to an embodiment, it may be assumed that at least one node to be checked (for example, a1 to a3) affects a connected node (for example, a third neighboring node u3) according to a predetermined probability p. , Determination of at least one node a2 and a3 determined to have an influence may be performed based on such a probability. The predetermined probability p may be predefined, may be selected within a predefined range, or may be selected from a plurality of predefined probability values. In the latter case, for example, the probability p may be uniformly selected from 0.01 and 0.01. As another example, the probability p may be randomly selected from 0.1, 0.01, and 0.001. In addition, the probability p can be variously defined according to the designer's choice.

, 예를 들어, a2 및 a3)가 결정되면, 현 활성 노드(

, a2 및 a3)를 기반으로 신규 활성 노드(

) 및 총 기여 노드(

)가 갱신된다. 구체적으로, 신규 활성 노드(

)는 현 활성 노드(

, a2 및 a3) 중에서 총 기여 노드(

)와 중첩되지 않는 노드(a2 및 a3)로 재결정되어 갱신되고, 총 기여 노드(

)는 기존의 총 기여 노드(

)에 현 활성 노드(

)를 부가하여 갱신된다. 다시 말해서, 신규 활성 노드(

)는 기존의 다른 처리 과정(1^st 및 두 번째 처리 과정)에서 검색된 노드들을 배제하고, 새롭게 활성화된 노드만을 다음 처리 과정(세 번째 처리 과정)으로 전달하도록 갱신될 수 있고, 총 기여 노드(

)는 기존의 다른 처리 과정(1^st 및 두 번째 처리 과정)에서 검색된 노드들을 모두 다음 처리 과정(세 번째 처리 과정)으로 전달하도록 갱신될 수 있다. 예를 들어, 신규 활성 노드(

) 및 총 기여 노드(

)가 {u3}으로 초기화된 경우, 신규 활성 노드(

)는 {a2, a3}로 갱신되고, 총 기여 노드(

)는 {u3, a2, a3}로 갱신된다.Currently active node(

, For example, a2 and a3) are determined, the currently active node (

, a2 and a3) based on the new active node (

) And the total contributing node (

) Is updated. Specifically, the new active node (

) Is the currently active node (

, a2 and a3) of the total contributing nodes (

) And the nodes that do not overlap (a2 and a3) are re-determined and updated, and the total contributing node (

) Is the existing total contributing node (

) To the currently active node (

It is updated by adding ). In other words, the new active node (

) It may be renewed excluding the retrieved nodes in the other existing process (1 ^st and the second process), and only the newly activated node to pass to the next process (the third process), the total contribution node (

) Can be updated all the retrieved from the other existing process (1 ^st and second processing) nodes to forward to the next process (the third process). For example, a new active node (

) And the total contributing node (

) Is initialized to {u3}, the new active node (

) Is updated to {a2, a3}, and the total contributing node (

) Is updated to {u3, a2, a3}.

) 및 총 기여 노드(

)의 갱신은 하기의 수학식 4 및 수학식 5로 각각 표현될 수 있다.New active node (

) And the total contributing node (

) Can be represented by Equation 4 and Equation 5 below.

[Equation 4]

[Equation 5]

5 is a diagram showing an embodiment of a node to be checked for a current active node.

)의 요소의 개수가 0이 될 때까지 반복될 수 있다(p3, 이하 제3 처리 과정). 다시 말해서,

을 만족할 때까지, 제2 처리 과정(두 번째 만약 신규 활성 노드(

)의 요소의 개수가 0이 아니라면, 도 5에 도시된 바와 같이, 신규 활성 노드(

)에 대해서 동일한 과정(p3)이 반복될 수 있다. 상세하게는 신규 활성 노드(

)인 적어도 하나의 노드(a2, a3) 각각에 대해서 영향을 미치는 적어도 하나의 노드(a3, b1, b3)를 검출하고, 검출한 적어도 하나의 노드(a3, b1, b3)로 현 활성 노드(

)로 분류하고, 현 활성 노드(

) 중에서 총 기여 노드(

)와 중첩되지 않는 노드(a2 및 a3)를 신규 활성 노드(

)로 다시 설정하고, 기존의 총 기여 노드(

, 즉 제2 처리 과정(p2)에서 연산된 총 기여 노드)에 현 활성 노드(

)를 부가하여 총 기여 노드(

)를 갱신할 수 있다.The second processing process p2 described above is the same or partially modified, and thus a new active node (

It may be repeated until the number of elements of) becomes 0 (p3, hereinafter, a third processing process). In other words,

Until it is satisfied, the second processing process (the second if a new active node (

If the number of elements of) is not 0, as shown in FIG. 5, a new active node (

), the same process (p3) may be repeated. Specifically, a new active node (

), at least one node (a3, b1, b3) affecting each of the at least one node (a2, a3) is detected, and the currently active node (a3, b1, b3) as the detected at least one node (

), and the currently active node (

) Of the total contributing nodes (

) And nodes (a2 and a3) that do not overlap with the newly active node (

), and the existing total contribution node (

In other words, the currently active node (the total contributing node calculated in the second processing process p2)

) To add the total contribution node (

) Can be updated.

)로 판단된 노드(a2, a3) 중에서 어느 하나의 노드(a2, 이하 제1 신규 활성 노드)의 경우, 제1 신규 활성 노드(a2)와 연결된 적어도 하나의 노드들(a3, b1, b2, b3)이 확인 대상 노드(

)로 결정되고, 상술한 바와 동일하게 확인 대상 노드(

)에 속하는 적어도 하나의 노드들(a3, b1, b2, b3) 중에서 현 활성 노드(

)가 검출될 수 있다. 예를 들어, 제1 노드(a2)와 연결된 노드(a3, b1 내지 b3) 중 활성화된 노드(a3, b1, b3)가 현 활성 노드(

)로 판단되어 검출된다. 여기서, 현 활성 노드(

)는 기존의 과정(p2)에서 현 활성 노드(

) 및/또는 신규 활성 노드(

)로 판단된 노드(일례로, a3)도 포함될 수 있다. 순차적으로 현 활성 노드(

) 중에서 기존 처리 과정(p2)의 총 기여 노드(

, 예를 들어,

={u3, a2, a3})와 중첩되지 않는 노드(b1 및 b3)가, 제1 신규 활성 노드(a2)에 대응하는 새로운 신규 활성 노드(

, 예를 들어,

={b1, b3})로 결정될 수 있다. 아울러 제2 처리 과정(p2)에서 연산된 총 기여 노드(

, 예를 들어,

={u3, a2, a3})에 현 활성 노드(예를 들어,

={a3, b1, b3))가 부가되여 총 기여 노드(

)가 갱신되어 다시 획득될 수 있다. 이 경우, 갱신되어 다시 획득된 총 기여 노드(

)는, 예를 들어, {u3, a2, a3, b1, b3}로 주어질 수 있다. For a more specific example, in the second processing process p2 described above, a new active node (

), in the case of any one node (a2, hereinafter referred to as the first new active node) among the nodes a2 and a3 determined to be, at least one node (a3, b1, b2) connected to the first newly active node (a2) b3) the node to be checked (

), and the node to be checked (

) Of the at least one node (a3, b1, b2, b3) belonging to the currently active node (

) Can be detected. For example, among the nodes a3, b1 to b3 connected to the first node a2, the active node a3, b1, b3 is the currently active node (

) Is determined and detected. Here, the currently active node (

) Is the currently active node (

) And/or a new active node (

A node determined as) (for example, a3) may also be included. Sequentially currently active nodes (

) Of the total contributing nodes of the existing process (p2) (

, For example,

={u3, a2, a3}) and nodes b1 and b3 that do not overlap each other are the new new active nodes corresponding to the first new active node a2 (

, For example,

={b1, b3}). In addition, the total contribution node calculated in the second processing process (p2) (

, For example,

={u3, a2, a3}) on the currently active node (for example,

=(a3, b1, b3)) is added to the total contributing node (

) Can be updated and acquired again. In this case, the total contributing nodes that were updated and reacquired (

) May be given as {u3, a2, a3, b1, b3}, for example.

, 예를 들어,

={b1, b3})는 적어도 하나의 요소(즉, 노드(b1, b3)를 가지고 있으므로, 새로운 신규 활성 노드(

)의 노드(b1, b3) 각각에 대해서도 상술한 제2 처리 과정(p2) 또는 제3 처리 과정(p3)과 동일한 과정이 다시 반복된다. 이에 따라 총 기여 노드(

)는 다시 갱신된다. 또 다른 과정의 반복에서도 신규 활성 노드(

)가 공집합이 아니라면, 동일하게 총 기여 노드(

)는 계속해서 갱신되게 된다.A new new active node corresponding to the first new active node a2 (

, For example,

={b1, b3}) has at least one element (i.e. nodes (b1, b3)), so a new new active node (

For each of the nodes b1 and b3 of ), the same process as the above-described second process p2 or third process p3 is repeated again. Accordingly, the total contributing node (

) Is updated again. In another iteration of the process, the new active node (

) Is not an empty set, then the total contributing node (

) Will be continuously updated.

)로 판단된 노드(a2, a3) 중에서 다른 하나의 노드(a3, 이하 제2 신규 활성 노드)의 경우, 확인 대상 노드(

)는 {a2}로 결정되고, 현 활성 노드(

)는 {a2}로 결정될 수 있다. 이때, 현 활성 노드(

) 내에는 제2 처리 과정(p2)에서 연산된 총 기여 노드(

, 예를 들어,

={u3, a2, a3})와 중첩되지 않는 노드가 부재하므로, 제2 신규 활성 노드(a3)에 대응하는 새로운 신규 활성 노드(

)는 어떠한 요소도 갖지 않는다(즉,

). 이와 같이 신규 활성 노드(

)가 공집합인 경우, 제2 신규 활성 노드(a3)에 대한 처리 과정(즉, 총 기여 노드(

)의 갱신 과정)은 종료된다. 한편, 현 활성 노드(

)의 모든 요소는, 상술한 제2 처리 과정(p2)에서 연산된 총 기여 노드(

)의 모든 요소와 동일하므로, 총 기여 노드(

)는 제2 처리 과정(p2)에서 연산된 총 기여 노드와 동일하게 된다(즉,

={u3, a2, a3}).In addition, in the second processing process p2 described above in the same example, a new active node (

) In the case of the other node (a3, hereinafter the second new active node) among the nodes a2 and a3 determined as

) Is determined by {a2}, and the currently active node (

) Can be determined as {a2}. At this time, the currently active node (

), the total contribution node calculated in the second processing process (p2) (

, For example,

={u3, a2, a3}) and there is no node that does not overlap, so a new new active node corresponding to the second new active node a3 (

) Has no elements (i.e.

). Like this, the new active node (

) Is the empty set, the processing process for the second new active node a3 (that is, the total contributing node (

) Update process) is terminated. Meanwhile, the currently active node (

All elements of) are the total contributing nodes calculated in the second processing process p2 described above (

), so the total contributing node (

) Becomes the same as the total contributing node calculated in the second processing process (p2) (i.e.

={u3, a2, a3}).

As described above, the second processing step p2 and the third processing step p3 described above may be performed on all activated neighbor nodes u1, u3, and u5. Also, depending on the situation, additional processing process(s) may be performed for all activated neighboring nodes u1, u3, and u5 even after the third processing process p3.

)가 획득되면, 총 기여 노드 결정부(122)는 획득된 총 기여 노드(

)를, 고영향 노드 획득부(124)로 전달할 수 있다. 이 경우, 총 기여 노드(

)는 전기적 신호의 형태로 전달될 수도 있다.Total Contributing Node(

) Is obtained, the total contributing node determination unit 122 obtained total contributing node (

) May be transmitted to the high-impact node acquisition unit 124. In this case, the total contributing node (

) May be transmitted in the form of an electrical signal.

)를 기반으로 목표 노드(u)에 대해 상대적으로 높은 영향력을 갖는 적어도 하나의 이웃 노드(이하 고 영향 노드)를 이웃 노드(u1 내지 u5)로부터 선택할 수 있다.According to an embodiment, the highly affected node acquisition unit 124 is a total contributing node for all or part of each of the neighboring nodes u1 to u5 (

Based on ), at least one neighboring node (hereinafter, a high influence node) having a relatively high influence on the target node u may be selected from the neighboring nodes u1 to u5.

, 휴리스틱 영향 가중치 등으로도 표현 가능하다)을 이용할 수도 있다. 휴리스틱 영향력(

)은, 그리디 최적화 기술과 더불어 역 선형 임계 모델에서 검출되는 노드의 개수를 최소화할 수 있도록 이용된다. 또한, 고영향 노드 획득부(124)는 역 선형 임계 모델을 이용하여 최적의 고 영향 노드를 결정할 수도 있다.According to an embodiment, the high-impact node acquisition unit 124 calculates a weight for the influence of each of the neighboring nodes (u1 to u5).

, It can also be expressed as a heuristic influence weight). Heuristic influence (

) Is used to minimize the number of nodes detected in the inverse linear threshold model together with the greedy optimization technique. In addition, the high-impact node acquisition unit 124 may determine an optimal high-impact node using an inverse linear threshold model.

)를 기반으로 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)를 연산하고, 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)을 정규화하여 정규화된 영향력, 즉 휴리스틱 영향력(

)를 획득한 후, 이를 이용하여 목표 노드(v)에 대응하는 적어도 하나의 이웃 노드(u1 내지 u5) 중에서 각 이웃 노드(u1 내지 u5)의 휴리스틱 영향력(

)의 총합(

)을 최대로 하는 적어도 하나의 고 영향 노드(u1 내지 u5 중 적어도 하나)를 획득할 수 있다.In more detail, the high-impact node acquisition unit 124 is a total contribution node (

) Based on the influence of one neighboring node (u) on the target node (v) (

), and the influence of one neighboring node (u) on the target node (v) (

) By normalizing the normalized influence, that is, the heuristic influence (

), and using this, the heuristic influence of each neighboring node (u1 to u5) among at least one neighboring node (u1 to u5) corresponding to the target node (v) (

) Of the sum (

At least one high-impact node (at least one of u1 to u5) that maximizes) may be acquired.

)은, 측정된 총 기여 노드(

)의 개수(

)와 일정한 관계를 갖는 것으로 설정될 수 있다. 예를 들어, 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)은, 총 기여 노드(

)의 개수(

)와 반비례 관계를 갖는 것으로 설정 및 정의될 수 있다. 이는 하기의 수학식 6으로 표현 가능하다.Specifically, the influence of one neighboring node (u) on the target node (v) (

) Is the measured total contribution node (

) Of the number (

) And can be set to have a certain relationship. For example, the influence of one neighboring node (u) on the target node (v) (

) Is the total contributing node (

) Of the number (

) And can be defined and defined as having an inverse relationship with each other. This can be expressed by Equation 6 below.

[Equation 6]

)과 측정된 총 기여 노드(

)의 개수(

) 사이의 관계는 반비례 관계 이외의 다른 관계를 갖는 것도 가능하다. 예를 들어, 영향력(

)과 총 기여 노드(

)의 개수(

) 사이의 관계는, 다른 함수(예를 들어, 음의 지수 함수나 음의 로그 함수) 등으로 설정될 수도 있다.Of course, depending on the designer, the influence of one neighboring node (u) on the target node (v) (

) And the measured total contribution node (

) Of the number (

It is also possible for the relationship between) to have a relationship other than the inverse relationship. For example, influence(

) And the total contributing node (

) Of the number (

The relationship between) may be set with another function (eg, a negative exponential function or a negative log function).

)은, 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)을 정규화하여 획득될 수 있다. 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)은 하기의 수학식 7과 같이 최소-최대 정리(Min-Max theorem)를 이용하여 정규화될 수 있다.Heuristic influence (

) Is the influence of any one neighboring node (u) on the target node (v) (

) Can be obtained by normalizing. The influence of one neighboring node (u) on the target node (v) (

) May be normalized using the Min-Max theorem as shown in Equation 7 below.

[Equation 7]

)은, 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)과 최소 영향력(

) 사이의 차이를, 최대 영향력(

)과 최소 영향력(

) 사이의 차이로 나눈 값(즉, 이들 사이의 비)으로 주어질 수 있다. 만약 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)이 최대 영향력(

)과 동일하면, 휴리스틱 영향력(

)은, 1의 값을 가지게 된다. 반대로 목표 노드(v)에 대한 어느 하나의 이웃 노드(u)의 영향력(

)이 최소 영향력(

)과 동일하다면 휴리스틱 영향력(

)은, 0의 값을 갖게 된다. 다시 말해서, 이와 같이 정규화된 영향력(

)은 최소 0에서 최대 1 사이의 값을 가지게 된다.Specifically normalized influence, that is, heuristic influence (

) Is the influence of any one neighboring node (u) on the target node (v) (

) And least influence (

), the maximum influence (

) And least influence (

) Divided by the difference between them (i.e. the ratio between them). If the influence of one neighboring node (u) on the target node (v) (

) Is the maximum influence (

) Equals the heuristic influence (

) Has a value of 1. Conversely, the influence of one neighboring node (u) on the target node (v) (

) Is the least influential (

) Equals heuristic influence (

) Has a value of 0. In other words, this normalized influence (

) Has a value between minimum 0 and maximum 1.

)의 개수(

=

)가 상대적으로 작은 노드는, 상대적으로 강한 영향력을 가지고 있을 확률이 더 높다는 것을 나타낸다. 이런 이유로 높은 휴리스틱 영향력(

)을 갖는 노드(들)를 선택하면, 목표 노드(v)에 충분한 영향을 끼칠 수 있으면서도 상대적으로 적은 개수 또는 최소한의 노드(들)을 획득할 수 있게 된다. 만약 목표 노드(v)가 소셜 네트워크의 목표 사용자(일례로 시드 사용자)라면, 적은 수의 사용자의 활성화만으로도 소셜 네트워크 상의 목표 사용자를 활성화시킬 수 있게 되고, 이에 따라 비용 절감 효과가 발생하게 된다.Referring to Equations 6 and 7, the total contribution node (

) Of the number (

=

A node with a relatively small) indicates that it is more likely to have a relatively strong influence. For this reason, high heuristic influence (

If the node(s) having) is selected, it is possible to obtain a relatively small number or minimum number of node(s) while sufficiently affecting the target node (v). If the target node v is a target user of the social network (for example, a seed user), the target user on the social network can be activated by only activating a small number of users, resulting in cost reduction.

)을 결정하기 위하여, 고영향 노드 획득부(124)는 역 선형 임계 모델을 이용한다. 역 선형 모델 내에서 최대 휴리스틱 영향력(

)을 갖는 이웃 노드(u)는 그리디 알고리즘을 이용하여 획득된다. 이에 따라 목표 노드(v)에 대한 최소 개수의 고 영향 노드로 이루어진 노드 집단(

)은, 하기의 수학식 8과 같이 주어진다.The group consisting of the minimum number of high impact nodes for the target node (v) (

), the high-impact node acquisition unit 124 uses an inverse linear threshold model. Maximum heuristic influence (

The neighboring node u with) is obtained using the greedy algorithm. Accordingly, a node group consisting of the minimum number of highly influential nodes for the target node (v) (

) Is given as in Equation 8 below.

[Equation 8]

)은, 목표 노드(v)에 대한 적어도 하나의 이웃 노드 u1 내지 u5) 중에서, 각각의 노드(u1 내지 u5 중 적어도 하나)의 휴리스틱 영향력(들) (

)의 총합(

)이 최대 값이 되도록 하는 적어도 하나의 고 영향 노드(u1 내지 u5 중 적어도 하나,

)의 집합일 수 있다. 다시 말해서, 상술한 노드 집단(

)은 고 영향 노드(u1 내지 u5 중 적어도 하나,

)들을 포함하며, 이들 고 영향 노드(들)(u1 내지 u5 중 적어도 하나) 각각의 휴리스틱 영향력(

)의 총합(

)은 최대가 된다.According to Equation 8, the above-described node group (

) Is the heuristic influence(s) of each node (at least one of u1 to u5) among at least one neighboring node u1 to u5 to the target node (v) (

) Of the sum (

) At least one high influence node (at least one of u1 to u5,

) Can be a set. In other words, the group of nodes described above (

) Is the high impact node (at least one of u1 to u5,

), and the heuristic influence of each of these high influence node(s) (at least one of u1 to u5) (

) Of the sum (

) Becomes maximum.

는 임계 값을 의미한다. 임계 값(

)은 선택된 고 영향 노드 각각의 휴리스틱 영향력(

)들의 합(

)과 비교될 수 있다. 만약 고 영향 노드 각각의 휴리스틱 영향력(

)들의 합(

)이 임계 값(

))보다 작다면, 후술하는 바와 같이 이웃 노드는 더 선택되고, 만약 고 영향 노드 각각의 휴리스틱 영향력(

)들의 합(

)이 임계 값(

)보다 크다면, 목표 노드(v)에 대한 고 영향 노드가 확정되게 된다. 한편, 수학식 7에 따르면, 가장 큰 휴리스틱 영향력(

)은 그 값이 1.000을 가지게 될 수 있으므로, 임계 값(

)은 선택될 고 영향 노드의 개수를 의미하게 될 수도 있다. Meanwhile, in Equation 8,

Means the threshold. Threshold(

) Is the heuristic influence of each selected high influence node (

) Sum (

) Can be compared. If the heuristic influence of each high-impact node (

) Sum (

) Is the threshold (

)), as described later, the neighboring node is more selected, and if the heuristic influence of each of the high impact nodes (

) Sum (

) Is the threshold (

If it is larger than ), the high influence node for the target node v is determined. Meanwhile, according to Equation 7, the largest heuristic influence (

) Can have a value of 1.000, so the threshold value (

) May mean the number of high influence nodes to be selected.

)는 다수 투표 기법(majority voting technique)에 의해서 설정될 수도 있다. 다시 말해서, 목표 노드(v)는, 이웃 노드(u1 내지 u5) 중 과반수의 노드의 활성화에 응하여 활성화되는 것으로 설정될 수도 있다. 이 경우, 임계 값(

)는 하기의 수학식 9으로 주어질 수 있다.According to one embodiment, the threshold value (

) May be established by a majority voting technique. In other words, the target node v may be set to be activated in response to activation of a majority of the neighboring nodes u1 to u5. In this case, the threshold (

) Can be given by Equation 9 below.

[Equation 9]

)은 항상 1.000을 가질 수 있으며, 이에 따라 수학식 9로 주어지는 임계 값(

)은 휴리스틱 영향력의 총합(

)에 대해 어떠한 문제도 야기하지 않는다.According to the method described above, when any one neighboring node u is selected, the heuristic influence corresponding to the selected neighboring node u (

) Can always have 1.000, and accordingly, the threshold value given by Equation 9 (

) Is the sum of the heuristic influences (

) Does not cause any problems.

)는, 하기의 도 10에 도시된 바와 같이, 적어도 하나의 이웃 노드(u)의 총 기여 노드(

)의 합집합으로 주어질 수 있다.Meanwhile, the total contributing node to the target node (v) (

) Is, as shown in FIG. 10 below, the total contributing node of at least one neighboring node u (

Can be given as the union of ).

[Equation 10]

)는, 하기의 수학식 11과 같이, 적어도 하나의 목표 노드(v)에 대한 총 기여 노드(

)의 합집합으로 주어진다.Total contributing nodes of the entire target node group S including at least one target node v

) Is a total contribution node for at least one target node (v) as shown in Equation 11 below (

It is given as the union of ).

[Equation 11]

)의 개수(

)는 하기의 수학식 12와 같이 연산될 수 있다.In addition, the total contributing nodes of the total target node group (S) (

) Of the number (

) Can be calculated as in Equation 12 below.

[Equation 12]

Accordingly, a minimum value of the minimum number of node(s) and node(s) required to affect all target node(s)(S) within one network 10 may be obtained. As described above, since the number of node(s) can be evaluated at cost, it is possible to activate the target node(s)(v) at a minimum cost, and also target node(s)() in one network 10 It is possible to determine the minimum cost required for activation of v).

)의 획득 과정을 보다 구체적으로 설명한다.Hereinafter, with reference to FIGS. 6 to 8, a group consisting of the minimum number of highly affected nodes for the target node v by the highly affected node acquisition unit 124 (

) Will be described in more detail.

), 특정한 이웃 노드(ui)에 대한 총 기여 노드(

), 총 기여 노드(

) 내의 노드의 개수(

), 영향력(

), 휴리스틱 영향력(

), 휴리스틱 영향력(

)에 따라 선택된 이웃 노드(u), 고 영향 노드로 이루어진 집단(

) 및 고 영향 노드로 이루어진 집단(

)에 대응하는 총 기여 노드(

)를 나타낸다. 도 7은 적어도 하나의 이웃 노드로부터 가장 높은 영향력의 이웃 노드를 결정하는 방법의 일 실시예를 도시한 제1 도이다. 도 8은 적어도 하나의 이웃 노드로부터 가장 높은 영향력의 이웃 노드를 결정하는 방법의 일 실시예를 도시한 제2 도이며, 도 9는 적어도 하나의 이웃 노드로부터 가장 높은 영향력의 이웃 노드를 결정하는 방법의 일 실시예를 도시한 제3 도이다. 도 6 내지 도 9는 임계 값(

)이 3인 경우를 가정하여 실행된 일례를 도시한 것이다.6 is a diagram illustrating an embodiment of a method of determining a neighboring node having the highest influence from at least one neighboring node. Each column of FIG. 6 has a threshold value (

), the total contributing node for a specific neighboring node (ui) (

), the total contributing node (

The number of nodes in (

), Influence(

), heuristic influence (

), heuristic influence (

), a group consisting of a neighboring node (u) selected according to the high influence node (

) And a group of high-impact nodes (

Total contributing nodes corresponding to) (

). 7 is a first diagram illustrating an embodiment of a method of determining a neighboring node having the highest influence from at least one neighboring node. FIG. 8 is a second diagram illustrating an embodiment of a method of determining a neighboring node having the highest influence from at least one neighboring node, and FIG. 9 is a method of determining a neighboring node having the highest influence from at least one neighboring node. 3 is a diagram showing an embodiment of. 6 to 9 show the threshold value (

) Shows an example executed assuming the case of 3.

,

,

,

및

)가 획득될 수 있다. 예를 들어, 도 7에 도시된 바와 같이, 첫 번째 반복 과정(I1)에서, 각 이웃 노드(u1 내지 u5)의 총 기여 노드에 해당하는 노드의 개수(

,

,

,

및

)는 각각 5, 6, 3, 7 및 4로 주어질 수 있다. 이 경우, 각 이웃 노드(u1 내지 u5)에 대응하는 영향력(

)은, 순차적으로 0.200, 0.167, 0.333, 0.143 및 0.250으로 주어지고, 이를 기반으로 연산된 휴리스틱 영향력(

)은 각각 0.300, 0.126, 1.000, 0.000 및 0.263으로 주어질 수 있다. 여기서, 제3 이웃 노드(u3)의 휴리스틱 영향력(

)이 가장 큰 값(즉, 1.000)을 가지므로, 이들 노드(u1 내지 u5) 중에서 제3 이웃 노드(u3)가 선택되게 된다. 이에 따라 고 영향 노드로 이루어진 집단(

)은 제3 이웃 노드(u3)를 포함하게 되며, 고 영향 노드로 이루어진 집단(

)에 대응하는 총 기여 노드(

)는, 제3 이웃 노드(u3)와 동일하게 {1, 4, 5}로 주어진다.As shown in FIG. 6, when a plurality of neighboring nodes u1 to u5 exist for each of at least one target node v, each of the total contributing nodes for each neighboring node u1 to u5 (

,

,

,

And

) Can be obtained. For example, as shown in FIG. 7, in the first iteration process I1, the number of nodes corresponding to the total contributing nodes of each of the neighboring nodes u1 to u5 (

,

,

,

And

) Can be given as 5, 6, 3, 7 and 4 respectively. In this case, the influence corresponding to each neighboring node (u1 to u5) (

) Is sequentially given as 0.200, 0.167, 0.333, 0.143 and 0.250, and the heuristic influence calculated based on this (

) Can be given as 0.300, 0.126, 1.000, 0.000 and 0.263 respectively. Here, the heuristic influence of the third neighboring node u3 (

) Has the largest value (that is, 1.000), the third neighboring node u3 is selected from among these nodes u1 to u5. Accordingly, a group consisting of high-impact nodes (

) Includes a third neighboring node (u3), and a group consisting of high-impact nodes (

Total contributing nodes corresponding to) (

) Is given by {1, 4, 5} in the same way as the third neighboring node u3.

)이 3으로 설정되었다면, 휴리스틱 영향력(

)의 총합(

)은 1.000이므로 휴리스틱 영향력(

)의 총합(

)은 임계 값(

)보다 작게 된다. 그러므로, 제2 반복 과정(I2)이 더 수행되게 된다. 이 경우, 제1 반복 과정(I3)에서 선택된 제3 이웃 노드(u3)에 대응하는 각각의 노드({1, 4, 5})는 나머지 다른 이웃 노드(u1, u2, u4, u5)에 부가될 수 있으며(예를 들어, 제3 이웃 노드(u3)에 대응하는 각각의 노드({1, 4, 5}와 나머지 다른 이웃 노드(u1, u2, u4, u5)에 대응하는 각각의 노드의 합집합이 획득될 수 있다), 나머지 다른 이웃 노드(u1, u2, u4, u5)에 대해서 상술한 바와 동일하게 제2 반복 과정(I2)이 수행될 수 있다. 이 경우, 각 노드(u1, u2, u4, u5)에 대응하는 휴리스틱 영향력(

)은, 예를 들어, 각각 1.000, 0.250, 0.000 및 0.430으로 주어질 수 있다. 여기서는, 제1 이웃 노드(u1)의 휴리스틱 영향력(

)이 가장 큰 값(즉, 1.000)을 가지므로, 나머지 다른 이웃 노드(u1, u2, u4, u5) 중에서 제1 이웃 노드(u1)가 선택되게 된다. 이에 따라 고 영향 노드로 이루어진 집단(

)은 제1 이웃 노드(u1) 및 제3 이웃 노드(u3)를 포함하게 되고, 고 영향 노드로 이루어진 집단(

)에 대한 총 기여 노드(

)는, 제1 이웃 노드(u1) 및 제3 이웃 노드(u3)의 합집합과 동일하게 {1, 2, 3, 4, 5}로 주어지게 된다.If threshold(

) Is set to 3, then heuristic influence (

) Of the sum (

) Is 1.000, so the heuristic influence (

) Of the sum (

) Is the threshold (

). Therefore, the second iteration process I2 is further performed. In this case, each node ({1, 4, 5}) corresponding to the third neighboring node u3 selected in the first iteration process (I3) is added to the other neighboring nodes (u1, u2, u4, u5). Can be (for example, each node corresponding to the third neighboring node u3 ({1, 4, 5}) and each node corresponding to the other neighboring nodes (u1, u2, u4, u5) A union can be obtained), and the second iterative process I2 may be performed on the remaining neighboring nodes u1, u2, u4, and u5 in the same manner as described above, in this case, each node u1 and u2. , u4, u5) heuristic influence (

) Can be given as, for example, 1.000, 0.250, 0.000 and 0.430, respectively. Here, the heuristic influence of the first neighboring node u1 (

) Has the largest value (i.e., 1.000), the first neighboring node u1 is selected from the other neighboring nodes u1, u2, u4, and u5. Accordingly, a group consisting of high-impact nodes (

) Includes the first neighboring node (u1) and the third neighboring node (u3), and a group consisting of high influence nodes (

) To the total contributing node (

) Is given by {1, 2, 3, 4, 5} in the same way as the union of the first and third neighboring nodes u1 and u3.

)의 총합(

)은 2.000으로, 여전히 휴리스틱 영향력(

)의 총합(

)이 임계 값(

)보다 작기 때문에, 제2 반복 과정(I2)에 이어 제3 반복 과정(I3)dl 수행되게 된다. 도 6 및 도 9에 도시된 바와 같이, 제3 반복 과정(I3)에서 제2 반복 과정(I2)에서 선택된 제1 이웃 노드(u1)의 각각의 노드({1,2,3,4,5})가 나머지 다른 이웃 노드(u2, u4, u5)에 추가될 수 있으며, 제1 이웃 노드(u1)의 각각의 노드({1,2,3,4,5})가 추가된 나머지 다른 이웃 노드(u2, u4, u5)에 대해서 상술한 과정(I1 또는 I2)이 반복 수행된다. 그 결과, 각 노드(u2, u4, u5)에 대응하는 휴리스틱 영향력(

)은, 예를 들어, 각각 0.667, 0.000 및 1.000으로 주어질 수 있으며, 동일하게 휴리스틱 영향력(

)이 가장 큰 제5 노드(u5)가 선택되게 된다. 따라서, 고 영향 노드로 이루어진 집단(

)은 제1 이웃 노드(u1), 제3 이웃 노드(u3) 및 제5 이웃 노드(u5)를 포함하여 갱신되고, 고 영향 노드로 이루어진 집단(

)에 대응하는 총 기여 노드(

)는, 제 이웃 노드(u1), 제3 이웃 노드(u3) 및 제5 이웃 노드(u5)들의 합집합과 동일하게 {1, 2, 3, 4, 5, 8, 15, 16}를 포함하게 된다. 최종적으로 총 8개의 노드(1, 2, 3, 4, 5, 8, 15, 16)를 포함하는 가장 영향력이 높은 노드로 선택되게 된다.Where heuristic influence (

) Of the sum (

) Is 2.000, still heuristic influence (

) Of the sum (

) Is the threshold (

), the third repetition process (I3) is performed following the second repetition process (I2). 6 and 9, each node ({1, 2, 3, 4, 5) of the first neighboring node u1 selected in the second iteration process I2 in the third iteration process I3 }) can be added to the remaining neighboring nodes (u2, u4, u5), and the remaining neighbors to which each node ({1,2,3,4,5}) of the first neighboring node (u1) is added. The above-described process (I1 or I2) is repeatedly performed for the nodes u2, u4, and u5. As a result, the heuristic influence corresponding to each node (u2, u4, u5) (

) Can be given as, for example, 0.667, 0.000 and 1.000, respectively, and equally heuristic influence (

The fifth node u5 with the largest) is selected. Therefore, a group of high-impact nodes (

) Is updated to include the first neighboring node (u1), the third neighboring node (u3), and the fifth neighboring node (u5), and a group consisting of high-impact nodes (

Total contributing nodes corresponding to) (

) Includes {1, 2, 3, 4, 5, 8, 15, 16} equal to the union of the third neighboring node u1, the third neighboring node u3, and the fifth neighboring node u5. do. Finally, it is selected as the most influential node including a total of 8 nodes (1, 2, 3, 4, 5, 8, 15, 16).

The processor 120 may be specially designed to perform the above-described operation, or may be specially programmed with a general information processing device for the above-described operation. Further, the processor 120 may be provided to be driven by calling a program for the above-described operation. In this case, the processor 120 may drive an application stored in a storage medium (not shown) to perform predefined command processing, calculation, determination, processing, and/or control operations. Here, the application stored in the storage medium may be previously created by a designer and stored in the storage medium, and/or may be acquired or updated through an electronic software distribution network accessible through a wired or wireless communication network.

The processor 120 is, for example, a central processing unit (CPU, Central Processing Unit), a microcontroller unit (MCU, Micro Controller Unit), a Micom (Micro Processor), an application processor (AP, Application Processor), electronic It may include a control unit (ECU, Electronic Controlling Unit) and/or other electronic devices capable of processing various operations and generating control signals. These devices can be implemented using, for example, one or more semiconductor chips and related components.

10 is a diagram showing an embodiment of an algorithm, specifically, the above-described roughly expressed as an algorithm.

)는 공집합으로 처리된다(line 1). 목표 노드(v)가 목표 노드 집단(S)에 속하는 경우에(line 2), 목표 노드(v)의 총 기여 노드(

)는 공집합으로 초기화되고(line 3), 신규 활성 노드(

)는 이웃 노드(u)를 포함하는 집합({u})으로 초기화된다(line 4). As shown in FIG. 10, first, when information (G(V, E)) on the network to be analyzed 10 and information on the group (S) of the target node (v) are input, the entire group (S) The total contribution node for (

) Is treated as an empty set (line 1). When the target node (v) belongs to the target node group (S) (line 2), the total contributing node of the target node (v) (

) Is initialized to the empty set (line 3), and the new active node (

) Is initialized to a set ({u}) containing the neighboring node (u) (line 4).

)에 해당하면, line 6 내지 line 14가 반복된다(line 5, line 15). The neighboring node (u) is the newly active node (

), lines 6 to 14 are repeated (line 5, line 15).

)에는 이웃 노드(u)와 관계있는 노드(들)의 집합(

)이 적용되고(line 6), 이웃 노드(u)에 대응하는 현 활성 노드(

)는 공집합으로 설정되며, 이웃 노드(u)에 대응하는 총 기여 노드(

)는 이웃 노드(u)를 포함하는 집합({u})으로 초기화된다(line 7).Specifically, the node to be checked (

) Is a set of node(s) related to the neighboring node (u) (

) Is applied (line 6), and the currently active node corresponding to the neighboring node (u) (

) Is set to the empty set, and the total contributing node corresponding to the neighboring node (u) (

) Is initialized to a set ({u}) containing the neighboring node (u) (line 7).

)에 해당하는 경우, 만약 이웃 노드(u)와 관계가 존재하는 노드(w)가 확률 p에 따라서 활성화되면, 현 활성 노드(

)와 {w}의 합집합으로 현 활성 노드(

)를 갱신하는 과정(line 9 내지 line 11)을 반복 수행한다(line 8 및 line 12).The node (w) that has a relationship with the neighboring node (u) is the node to be checked (

), if a node (w) having a relationship with a neighboring node (u) is activated according to the probability p, the currently active node (

) And {w} as the union of the currently active node (

) Is repeated (line 9 to line 11) (line 8 and line 12).

)는, 기존의 신규 활성 노드(

)에 이웃 노드(u)에 대한 총 기여 노드(

)와 목표 노드 집단(S)에 대한 총 기여 노드(

)를 차감하여 갱신하고(line 13), 아울러 이웃 노드(u)에 대한 총 기여 노드(

)는, 기존의 이웃 노드(u)에 대한 총 기여 노드(

)와 신규 활성 노드(

)의 합하여(즉, 합집합을 이용하여) 갱신한다(line 14).After the iteration (line 8 to line 12), the new active node (

), the existing new active node (

) To the total contributing node (

) And the total contributing nodes to the target node group (S) (

) To be updated (line 13), and the total contributing node to the neighboring node (u) (

) Is the total contributing node to the existing neighboring node (u) (

) And the new active node (

) By summation (i.e., using the union) (line 14).

)을 갖는 이웃 노드(u)의 가장 작은 집합을, 고 영향 노드로 이루어진 노드 집단(

)에 대입하고(line 16), 노드 집단(

)에 속하는 노드(들)(y)의 총 기여 노드(

)를 이용하여 목표 노드(v)에 대한 총 기여 노드(

)를 갱신한다(line 17). 아울러 전체 집단(S)에 대한 총 기여 노드(

)를, 전체 집단(S)에 대한 총 기여 노드(

)에 목표 노드(v)에 대한 총 기여 노드(

)를 합하여 갱신시킨다(line 18).When the above-described repetition (line 5 to line 15) ends, the maximum heuristic influence (

The smallest set of neighboring nodes (u) with ), and the node group consisting of high influence nodes (

) Into (line 16), and a group of nodes (

) Of the node(s) belonging to (y)

) To the total contributing node (

) (Line 17). In addition, the total contributing node for the entire group (S) (

), the total contributing node for the entire population (S) (

) To the target node (v), the total contribution node (

) And update (line 18).

)에 속하는 요소들의 총 개수(λ(S))가 획득되고(line 20), 획득된 총 개수(λ(S))는 외부로 출력된다(line 21).When this process is over (line 19), the total contributing node for the entire group (S) (

The total number (λ(S)) of elements belonging to) is obtained (line 20), and the acquired total number (λ(S)) is output to the outside (line 21).

The above-described algorithm may be processed by the above-described processor 120. The processor 120 may detect node(s) having a high influence on the target node in the predetermined group S according to the above-described algorithm and determine the number of them.

The above-described apparatus for determining a high-impact node may be used to minimize marketing costs in viral marketing according to an embodiment.

11 is a diagram showing an embodiment of a system for minimizing marketing costs in viral marketing, and FIG. 12 is a diagram showing an embodiment of a network in which viral marketing is performed.

As shown in FIG. 11, the marketing cost minimization system 1 in viral marketing includes the device 200 for determining a high impact node and at least one social network service providing system 300 (301, 302) accessible. The social network service providing system 300 (301, 302), the social network service providing system (300: 301, 302), and at least one terminal device 310 to be connected may be included.

As shown in FIG. 12, at least one social network service providing system 300 (301, 302) may include at least one user (z0 to z24) and a relationship between these users.

Here, at least one user may include a virtual user. The virtual user may include a user account, and/or may include a homepage, a blog, a micro blog, and/or a post, etc. accompanying it.

At least one user may include the target user(s) z0. The target user z0 refers to a user who can have a strong influence on a number of users in a social network service. For example, the target user z0 may include a celebrity, a celebrity company, or a virtual user representing them. At least one neighboring user z11 to z24 is a user who has a predetermined relationship directly or indirectly with the target user z0, for example, first neighboring users z11 to z13 or second neighboring users z21 to z24 It may include. For example, at least one first neighbor user (z11 to z13) is a friend of the target user (z0), is a neighbor, is directly or indirectly following the target user (z0), or is followed by the target user (z0). It may be surging. In addition, at least one neighboring user z11 to z24, as well as at least one first neighboring user z11 to z13 having a direct relationship with the target user z0, as well as at least one first neighboring user z11 to z At least one second neighboring user z21 to z24 directly or indirectly having a relationship with z13) or via other user(s) may also be included.

According to an embodiment, the marketing cost minimization system 1 may also include a plurality of social network service providing systems 301 and 302. For example, a plurality of social network service providing systems 301 and 302 may interwork with each other or may be provided to transmit and receive mutually necessary data. In this case, the apparatus 200 for determining a high-impact node may obtain information on a user and information on a relationship between users from each of the plurality of social network service providing systems 301 and 302.

The plurality of social network service providing systems 301 and 302 may be implemented using one or more computing devices, and the computing device may include a server computing device, a desktop computer, and/or various portable electronic devices. .

The high-influence node determination device 200 includes information on a target user on a social network from at least one social network service providing system 300 (301, 302) (for example, account information, friends, neighbors, followers/following). Information, etc.) and information about at least one neighboring user(s) corresponding to the target user may be obtained.

The apparatus 200 for determining a high-influence node, based on information obtained from at least one social network service providing system 300 (301, 302), has a high-influence user(s) having a high influence on the target user as described above. It is possible to obtain a group for and also obtain a cost based on the number of high impact user(s) in the group. More specifically, the apparatus 200 for determining a high impact node may correspond to at least one first neighboring user (z11 to z13, which may correspond to the above-described target node) corresponding to the target user (z0, which may correspond to the above-described target node). And the target user z0 and the neighboring users z11 to z24 directly or indirectly connected to the target user z0, and information on the relationship between them (z0 to z24). In addition, after determining at least one total contributing user (which may correspond to the above-described total contributing node) for each of the at least one first neighboring user (z11 to z13), the apparatus 200 for determining the high-influence node It may be designed to determine the neighboring user (which may correspond to the above-described high-influence node) with the highest influence on the target user among at least one neighboring user based on the user.

)의 크기(λ(S))를 연산하는 방법과 동일하거나 또는 일부 변형된 방법을 이용하여 수행될 수 있다. 고 영향 노드 결정 장치(200)의 동작 및 기능에 대한 자세한 내용은 도 1 내지 도 10을 통하여 기 설명하였으므로 중복 설명의 회피를 위해 이하 생략하도록 한다.In this case, the cost to be invested for activation of the target user can be determined by the total number of contributing users corresponding to the group consisting of the target users, which corresponds to the group (S) consisting of the target node (v) described above. Total Contributing Node(

The same as the method of calculating the size (λ(S)) of) or may be performed using a partially modified method. Details of the operation and function of the apparatus 200 for determining a high-influence node have been previously described with reference to FIGS. 1 to 10, and thus will be omitted in order to avoid redundant descriptions.

The terminal device 310 is provided to be able to access at least one social network service providing system 300 (301, 302) through a wired or wireless communication network. The terminal device 310 is configured to allow a user to access another user's home page, view, duplicate, share, cite, and/or transmit another user's post, and/or the user's homepage of the terminal device 310 or It may be used to view or edit a micro blog, or to create, view, modify, and/or delete a post or the like by a user of the terminal device 310.

The terminal device 310 is, for example, a smart phone, a tablet PC, a head mounted display (HMD) device, a smart watch, a digital television, a set-top box, a desktop computer, a laptop computer, a navigation device, and a personal digital assistant. (PDA, Personal Digital Assistant), a portable game console, an electronic blackboard, an electronic billboard, or at least one of various devices capable of inputting and modifying other codes.

In the above, the apparatus for determining a high-impact node and a system for minimizing marketing costs, which is an embodiment to which it is applied, have been described.

The optimization method used in the above-described heuristic mixing model to obtain high-impact nodes or to solve the minimization of marketing costs is a technique related to the Knapsack problem, and the heuristic mixing model is a neighboring node ( Neighbor Users). Furthermore, the inverse independent cascade model used in the node activation process is a variation of the traditional independent cascade, and the influence maximization problem under this traditional independent cascade model is an NP-hard problem. In addition, since the backpack problem is also NP-difficulty, the method of acquiring high-impact nodes or minimizing marketing costs under a heuristic hybrid model corresponds to NP-hard.

In addition, the above-described heuristic mixed model algorithm may be a 2 approximation algorithm such as Equation 13 below.

[Equation 13]

The above-described knapsack problem is a 2 approximation algorithm, and is a variation of the knapsack problem of the heuristic mixing model, so the heuristic mixing model represents the same approximation ratio as shown in Equation 14 below.

[Equation 14]

On the other hand, in relation to the derivation of the execution time, if the value of the average in-degree in the network G is d, the execution time of the heuristic mixing algorithm may be given by Equation 15 below.

[Equation 15]

Hereinafter, an embodiment of a method for determining a high influencing node will be described with reference to FIGS. 13 to 15.

13 is a first flowchart for explaining an embodiment of a method for determining a high impact node.

As shown in FIG. 13, first, at least one neighboring node corresponding to at least one target node may be determined (S100). If there are a plurality of target nodes, at least one neighboring node may be independently determined for each of the plurality of target nodes.

When at least one neighboring node corresponding to the target node is determined, at least one total contributing node may be determined corresponding to each of the at least one neighboring node (S110). According to an embodiment, the total contribution node may be obtained using an inverse independent cascade model.

When at least one total contributing node corresponding to each of the at least one neighboring node is obtained, based on the total contributing node, the most influential neighboring node (i.e., high influence node) among at least one neighboring node corresponding to the target node is It can be determined (S120). According to an embodiment, the determination of the neighboring node having the highest influence may be performed by combining at least two of an inverse linear threshold model, a heuristic technique, and a greedy optimization technique.

According to such a method, a high impact node for a target node can be determined, and the number of nodes corresponding to the high impact node can also be determined.

14 is a second flowchart for explaining an embodiment of a method for determining a high-impact node, illustrating a process in which a total contributing node is determined in more detail.

Referring to FIG. 14, in order to determine a total contributing node corresponding to a neighboring node, a new active node and a total contributing node corresponding to each of one neighboring node or two or more neighboring nodes are first initialized (S111). For example, a new active node and a total contributing node may each be initialized to include a corresponding neighboring node.

At the same time or sequentially, one neighboring node or a node to be checked for each of two or more neighboring nodes is determined (S112). The node to be checked may include at least one node that has a relationship with the neighboring node and may affect the neighboring node, and may include all nodes thus prepared according to embodiments.

When the node to be checked is determined, a currently active node may be determined among nodes to be checked in response thereto (S112). The currently active node may be obtained from among nodes to be checked based on a predetermined probability. The predetermined probability may be, for example, uniformly selected from 0.01 and 0.01, or may be arbitrarily selected from 0.1, 0.01 and 0.001.

When the currently active node is determined, a new active node and a total contributing node may be updated based on the currently active node (S114). Specifically, the new active node is updated to a node that does not overlap with the existing total contributing nodes (ie, total contributing nodes before update) among the currently active nodes (ie, the difference between the currently active nodes and the total contributing nodes). The total contributing node may be updated by adding the current active node to the existing total contributing node (the total contributing node may be updated by adding the new active node to the total contributing node). In other words, the total contributing node may be updated with the union of the existing total contributing node and the currently active node.

If the number of nodes belonging to the new active node is 0, it may be determined (S115). In other words, it may be determined whether the new active node is an empty set.

If the number of new active nodes is 0 (YES in S115), the total contributing nodes updated in step S114 described above are determined as the final total contributing nodes (S116).

Conversely, if the number of new active nodes is not 0 (No in S115, that is, if the new active node is not an empty set), at least one node to be checked for each of the at least one newly active node is determined (S117). Here, the node to be checked includes nodes that have a relationship with the newly active node.

A current active node may be sequentially determined for a node to be checked for a new active node (S113), and a new active node may be newly updated in the same manner as described above based on the determined current active node (S114). Likewise, the total contributing nodes may be newly updated using the newly determined active node in the same manner as described above.

Subsequently, whether the number of nodes belonging to the newly updated new active node is 0 may be determined equally (S115), and if the number of nodes belonging to the newly updated new active node is 0 (Yes of S115), If the newly updated total contributing node is determined as the final total contributing node (S116), and the number of nodes belonging to the newly updated new active node is not 0 (No in S115), the newly updated new active node The node to be checked for is determined again (S117). In other words, steps S112 (S117), S113, and S114 described above may be continuously repeated until the new active node becomes an empty set (S115).

Accordingly, a total contributing node to at least one neighboring node may be determined.

FIG. 15 is a third flowchart for explaining an embodiment of a method for determining a high-influence node, and illustrates a process of selecting a neighboring node with the highest influential power in more detail.

As shown in FIG. 15, when at least one total contributing node corresponding to each of at least one neighboring node is determined as shown in FIG. 14, each of the at least one neighboring node is A corresponding influence may be calculated (S121). Here, the influence may mean the influence of a neighboring node on the target node.

When the influence corresponding to each neighboring node is calculated, the influence is normalized (S122). Accordingly, a normalized influence (for example, a heuristic influence) corresponding to each influence may be obtained. The normalized influence may be calculated using Equation 7 or the like described above.

When each normalized influence corresponding to each neighboring node is obtained, each normalized influence is compared, and a neighbor node having the largest normalized influence value is detected according to the comparison result (S123). For example, a neighboring node given a normalized influence value of 1.000 may be detected.

A sum (k) of normalized influences corresponding to the sequentially detected neighboring nodes may be calculated (S124). When the neighboring node is first detected, the value of the normalized influence of the detected neighboring node may be given as the sum of the normalized influences (k).

The sum of the normalized influences (k) may be compared with a predefined threshold value (θ) (S125). If the sum of the normalized influences k is equal to the predefined threshold value θ (YES in S125), the neighboring node detected in step S123 described above may be determined as a high influence neighboring node (S126). In this case, the set of at least one node(s) directly or indirectly related to the high-impact neighboring node may be determined as the total contributing node to the target node, and the number of nodes belonging to the total contributing node may also be calculated. .

Conversely, if the sum of the normalized influences (k) is less than the predefined threshold value (θ) (No in S125), the detected neighboring node is replaced with another non-detected neighboring node (that is, the normalized influence is relatively small). A neighbor node) and may be merged (S127). In other words, the union of each of at least one other neighboring node and the detected neighboring node may be obtained.

Subsequently, the merged and newly acquired influence on the neighboring node and the normalized influence may be calculated again (S128). When the normalized influence on the newly acquired neighboring node is obtained, the normalized influences are compared with each other as described above, and at least one neighboring node among newly obtained neighboring nodes may be detected according to the comparison result ( S123). The detected neighboring nodes include neighboring nodes having the largest normalized influence value among newly acquired neighboring nodes.

Sequentially, the sum of the normalized influences (k) corresponding to the neighboring nodes detected in the above-described process (i.e., the first detected neighboring node and the newly detected neighboring node) is calculated (S124), and if the normalized sum of influences If (k) is equal to the predefined threshold value (θ) (example of S125), the neighboring node obtained in the above-described steps S121 to S123 and the neighboring node obtained through the above-described steps S127, S128, and S123 are highly affected. It may be determined as a neighboring node (S126).

Conversely, if the sum of the normalized influences (k) is less than the predefined threshold value (θ) (No in S125), the above-described steps S127 and S128 are repeated, and other neighboring nodes that are not detected are merged with the detected neighboring nodes. (S127), the corresponding influence and the normalized influence may be calculated again (S128). Subsequently, at least one neighboring node may be detected from among newly acquired neighboring nodes according to a result of comparison of the recalculated normalized influence (S123). Thereafter, the sum of the normalized influences k is calculated again (S124), and whether it is repeated may be determined according to the result of the calculation (S125).

High impact neighbor nodes can be determined through such a process.

Hereinafter, an embodiment of a method for determining the minimization of marketing costs will be described with reference to FIG. 16.

16 is a flowchart illustrating an embodiment of a method for determining the minimization of marketing costs.

As shown in FIG. 16, in the method for determining the minimization of marketing costs, at least one neighboring user corresponding to the at least one target user may first be determined (S400). In this case, if it is desired to obtain marketing costs for a plurality of target users, at least one neighboring user may be independently determined for each of the plurality of target users.

When at least one neighboring user corresponding to the target user is determined, at least one total contributing user corresponding to each of the at least one neighboring user may be sequentially determined (S410). As described above, the total contributing users may be acquired using the inverse independent cascade model.

When at least one total contributing user corresponding to each of the at least one neighboring user is acquired, a user with the highest influence (for example, a neighboring user) among users related to the target user may be determined based on the total contributing user (S420). . Likewise, the decision of the user with high influence can be performed by combining at least two of an inverse linear threshold model, a heuristic technique, and a greedy optimization technique.

According to this method, a user group that has a high influence on the target user can be obtained, and the marketing cost can be determined based on the number of users belonging to the user group. As described above, since a minimum required group of users is obtained for activation of target users having strong influence on various users, it is possible to determine necessary marketing costs, and to minimize and optimize necessary marketing costs.

At least one of the method of determining a high-influence node and a method of minimizing marketing costs according to the above-described embodiment may be implemented in the form of a program that can be driven by a computer device. Here, the program may include a program command, a data file, a data structure, or the like alone or in combination. The program may be designed and produced using machine code or high-level language code. The program may be specially designed to implement the above-described method, or may be implemented using various functions or definitions that are known and available to those skilled in the computer software field. In addition, here, the computer device may be implemented including a processor or a memory that enables the function of a program to be realized, and may further include a communication device if necessary.

A program for implementing at least one of the above-described method for determining a high-impact node and a method for determining marketing cost minimization may be recorded in a computer-readable recording medium. The computer-readable recording medium includes, for example, a magnetic disk storage medium such as a hard disk or a floppy disk, a magnetic tape, an optical recording medium such as a compact disk or a DVD, a magnetic-optical recording medium such as a floppy disk, and a ROM. , A semiconductor storage device such as RAM or flash memory, etc., may include various types of hardware devices capable of storing a specific program executed according to a call from a computer.

In the above, a method of determining a high-impact node, a device for determining a high-impact node, a method for minimizing marketing costs in viral marketing, and various embodiments of a system for performing this have been described. The apparatus, a method for minimizing marketing costs in viral marketing, and a system for performing the same are not limited only to the above-described embodiments. Various devices or methods that can be implemented by modifying and modifying based on the above-described embodiment by a person of ordinary skill in the relevant technical field are also the above-described method of determining the high-impact node, the device for determining the high-impact node, and marketing costs in viral marketing. It may be an example of a minimization method and a system for performing it. For example, the described techniques are performed in an order different from the described method, and/or components such as systems, structures, devices, circuits, etc. described are combined or combined in a form different from the described method, or other components or Even if it is replaced or replaced by an equivalent, it may be an embodiment of the above-described method for determining a high-impact node, a device for determining a high-affect node, a method for minimizing marketing costs in viral marketing, and a system for performing this.

1: Marketing cost minimization system
100: high impact node determination device
110: information collection unit 120: processor
122: total contributing node determination unit 124: neighbor node acquisition unit
190: node information providing device
200: Marketing cost minimization device
300: social network service device 310: terminal device

Claims (20)

As a method of determining a high-impact node performed by a determining device of a high-impact node,
Determining at least one neighboring node corresponding to the target node;
Determining at least one total contributing node corresponding to each of the at least one neighboring node; And
Determining a node with the highest influence from the at least one neighboring node based on the total contributing node; Including,
Determining the node with the highest influence from the at least one neighboring node based on the total contributing node,
Including, each obtaining the influence of the at least one neighboring node on the target node,
The influence of the at least one neighboring node is determined as an inverse number of total contributing nodes corresponding to the neighboring node,
Determining a group of neighboring nodes based on the total contributing nodes; further comprising, determining a node having the most influence in the group of neighboring nodes,
And obtaining at least one normalized influence by normalizing the at least one influence. The method of claim 1,
Determining at least one total contributing node corresponding to each of the at least one neighboring node,
Determining at least one first verification target node corresponding to the neighboring node;
Selecting at least one first currently active node affecting the at least one neighboring node from the at least one first verification target node;
And updating a contributing node corresponding to the neighboring node by adding the at least one first currently active node to at least one contributing node corresponding to the at least one neighboring node. The method of claim 2,
Selecting at least one first currently active node affecting the at least one neighboring node from the at least one first identification target node,
And selecting at least one first active node from among the at least one first identification target node based on a predefined probability. The method of claim 2,
Determining at least one total contributing node corresponding to each of the at least one neighboring node,
Searching for a first new active node among the at least one first currently active node, wherein the first newly active node includes a node that does not overlap with the contributing node. The method of claim 4,
Determining at least one total contributing node corresponding to each of the at least one neighboring node,
If the first new active node is not found among the at least one first currently active node, determining the contributing node as the total contributing node. The method of claim 5,
Determining at least one total contributing node corresponding to each of the at least one neighboring node,
Determining at least one second verification target node corresponding to each of the searched first new active nodes;
Selecting at least one second currently active node affecting the first new active node from the at least one second verification target node;
Updating a contributing node corresponding to the neighboring node by adding the at least one second currently active node to the at least one contributing node; And
Searching for a second new active node among the at least one second currently active node. The method of claim 4,
Initializing the contributing node; further comprising,
The contributing node is set and initialized to be the same as the first new active node. delete delete The method of claim 1,
Normalizing the at least one influence to obtain at least one normalized influence,
And normalizing the influence of the neighboring node on the target node by calculating a difference between the influence of the neighboring node and the minimum influence of the target node and the difference between the maximum influence and the minimum influence of the target node. How to determine the influence node. delete As a method of determining a high-influence node performed by a determination device of a high-influence node,
Determining at least one neighboring node corresponding to the target node;
Determining at least one total contributing node corresponding to each of the at least one neighboring node; And
Determining a node with the highest influence from the at least one neighboring node based on the total contributing node; Including,
Determining the node with the highest influence from the at least one neighboring node based on the total contributing node,
Calculating at least one influence on each of the at least one neighboring node corresponding to the target node;
Obtaining at least one normalized influence for the at least one influence; And
Determining a neighboring node having the highest influence by determining a normalized influence having the largest value among the at least one normalized influence, and detecting a neighbor node corresponding to the determined normalized influence;
The value of the influence is calculated by obtaining the inverse of the contributing node. The method of claim 12,
The step of determining a neighboring node with the highest influence from the at least one neighboring node based on the total contributing node,
Detecting a neighboring node having the highest influence until the sum of the at least one normalized influence is equal to or greater than a predetermined threshold value. delete delete delete delete delete delete delete

Images