sna-node-ties

Similar documents
300 구보학보 12집. 1),,.,,, TV,,.,,,,,,..,...,....,... (recall). 2) 1) 양웅, 김충현, 김태원, 광고표현 수사법에 따른 이해와 선호 효과: 브랜드 인지도와 의미고정의 영향을 중심으로, 광고학연구 18권 2호, 2007 여름


Microsoft PowerPoint - ch03ysk2012.ppt [호환 모드]

Vol.259 C O N T E N T S M O N T H L Y P U B L I C F I N A N C E F O R U M

Page 2 of 6 Here are the rules for conjugating Whether (or not) and If when using a Descriptive Verb. The only difference here from Action Verbs is wh


Page 2 of 5 아니다 means to not be, and is therefore the opposite of 이다. While English simply turns words like to be or to exist negative by adding not,

- 2 -

6자료집최종(6.8))

본문01

DBPIA-NURIMEDIA

歯1.PDF

Microsoft PowerPoint - 27.pptx


Vol.258 C O N T E N T S M O N T H L Y P U B L I C F I N A N C E F O R U M

04 형사판례연구 hwp

(JBE Vol. 21, No. 1, January 2016) (Regular Paper) 21 1, (JBE Vol. 21, No. 1, January 2016) ISSN 228

°í¼®ÁÖ Ãâ·Â

Journal of Educational Innovation Research 2018, Vol. 28, No. 1, pp DOI: * A Analysis of

DBPIA-NURIMEDIA



Á¶´öÈñ_0304_final.hwp

Microsoft PowerPoint - Freebairn, John_ppt

Microsoft PowerPoint - 7-Work and Energy.ppt

step 1-1

(Exposure) Exposure (Exposure Assesment) EMF Unknown to mechanism Health Effect (Effect) Unknown to mechanism Behavior pattern (Micro- Environment) Re

하나님의 선한 손의 도우심 이세상에서 가장 큰 축복은 하나님이 나와 함께 하시는 것입니다. 그 이 유는 하나님이 모든 축복의 근원이시기 때문입니다. 에스라서에 보면 하나님의 선한 손의 도우심이 함께 했던 사람의 이야기 가 나와 있는데 에스라 7장은 거듭해서 그 비결을

슬라이드 제목 없음

~41-기술2-충적지반


Hi-MO 애프터케어 시스템 편 5. 오비맥주 카스 카스 후레쉬 테이블 맥주는 천연식품이다 편 처음 스타일 그대로, 부탁 케어~ Hi-MO 애프터케어 시스템 지속적인 모발 관리로 끝까지 스타일이 유지되도록 독보적이다! 근데 그거 아세요? 맥주도 인공첨가물이

,,,.,,,, (, 2013).,.,, (,, 2011). (, 2007;, 2008), (, 2005;,, 2007).,, (,, 2010;, 2010), (2012),,,.. (, 2011:,, 2012). (2007) 26%., (,,, 2011;, 2006;


02이용배(239~253)ok

12È«±â¼±¿Ü339~370

pdf 16..


0125_ 워크샵 발표자료_완성.key

Journal of Educational Innovation Research 2018, Vol. 28, No. 4, pp DOI: * A Research Trend

<32352D342D313020C0D3C7FDBCB120C0E5C5C2BFEC D E687770>

Can032.hwp


untitled

<B1B9C1A6B0B3B9DFC7F9B7C25FC3B9B0C9C0BD5FB3BBC1F62E706466>

지능정보연구제 16 권제 1 호 2010 년 3 월 (pp.71~92),.,.,., Support Vector Machines,,., KOSPI200.,. * 지능정보연구제 16 권제 1 호 2010 년 3 월

<30322D28C6AF29C0CCB1E2B4EB35362D312E687770>

大学4年生の正社員内定要因に関する実証分析

서론 34 2

2 / 26

< C6AFC1FD28B0F1C7C1292E687770>

182 동북아역사논총 42호 금융정책이 조선에 어떤 영향을 미쳤는지를 살펴보고자 한다. 일제 대외금융 정책의 기본원칙은 각 식민지와 점령지마다 별도의 발권은행을 수립하여 일본 은행권이 아닌 각 지역 통화를 발행케 한 점에 있다. 이들 통화는 일본은행권 과 等 價 로 연

<31325FB1E8B0E6BCBA2E687770>

ePapyrus PDF Document

Journal of Educational Innovation Research 2018, Vol. 28, No. 3, pp.1-25 DOI: * An Analysis on Content

도비라

Vol.257 C O N T E N T S M O N T H L Y P U B L I C F I N A N C E F O R U M

민속지_이건욱T 최종

<31342D3034C0E5C7FDBFB52E687770>

<28BCF6BDC D B0E6B1E2B5B520C1F6BFAABAB020BFA9BCBAC0CFC0DAB8AE20C1A4C3A520C3DFC1F8C0FCB7AB5FC3D6C1BE E E687770>

43-4대지07한주성ok

UPMLOPEKAUWE.hwp

April 2014 BROWN Education Webzine vol.2 생명을 꿈꾸다 목차 From Editor 아침에는 다리가 4개,점심에는 2개, 저녁에는 3개인 것은? Guidance 익숙해지는 일상 속에서 우리아이 자립심 키우기 환경을 지키는 아이들의 좋은 습

44-4대지.07이영희532~

Chap 6: Graphs

High Resolution Disparity Map Generation Using TOF Depth Camera In this paper, we propose a high-resolution disparity map generation method using a lo

¹Ìµå¹Ì3Â÷Àμâ

Microsoft PowerPoint - ch07ysk2012.ppt [호환 모드]

<303720C7CFC1A4BCF86F6B2E687770>

#Ȳ¿ë¼®

:,,.,. 456, 253 ( 89, 164 ), 203 ( 44, 159 ). Cronbach α= ,.,,..,,,.,. :,, ( )

ÀÌÁÖÈñ.hwp

IKC43_06.hwp

<303220C7D1C5C2B9AE2E687770>

<C0C7B7CAC0C720BBE7C8B8C0FB20B1E2B4C9B0FA20BAAFC8AD5FC0CCC7F6BCDB2E687770>

유선종 문희명 정희남 - 베이비붐세대 소유 부동산의 강제매각 결정요인 분석.hwp

Journal of Educational Innovation Research 2019, Vol. 29, No. 1, pp DOI: : * Research Subject

I&IRC5 TG_08권

Buy one get one with discount promotional strategy

<3130C0E5>

112초등정답3-수학(01~16)ok

<B3EDB9AEC1FD5F3235C1FD2E687770>


< FC1A4BAB8B9FDC7D D325FC3D6C1BEBABB2E687770>

공급 에는 3권역 내에 준공된 프라임 오피스가 없었다. 4분기에는 3개동의 프라임 오피스가 신규로 준공 될 예정이다.(사옥1개동, 임대용 오피스 2개동) 수요와 공실률 2014년 10월 한국은행이 발표한 자료에 따르면 한국의 2014년 경제성장률 예측치는 3.5%로 지

<BFA9BAD02DB0A1BBF3B1A4B0ED28C0CCBCF6B9FC2920B3BBC1F62E706466>

대한한의학원전학회지26권4호-교정본(1125).hwp

<C7D1B9CEC1B7BEEEB9AEC7D03631C1FD28C3D6C1BE292E687770>

중견국외교연구회

우리들이 일반적으로 기호

DBPIA-NURIMEDIA

Let G = (V, E) be a connected, undirected graph with a real-valued weight function w defined on E. Let A be a set of E, possibly empty, that is includ

Microsoft Word - Westpac Korean Handouts.doc

삼교-1-4.hwp

< BFCFB7E15FC7D1B1B9C1A4BAB8B9FDC7D0C8B85F31352D31BCF6C1A4C8AEC0CE2E687770>

DBPIA-NURIMEDIA

(2005) ,,.,..,,..,.,,,,,

<32382DC3BBB0A2C0E5BED6C0DA2E687770>


[ReadyToCameral]RUF¹öÆÛ(CSTA02-29).hwp

Transcription:

Node Centrality in Social Networks Nov. 2015 Youn-Hee Han http://link.koreatech.ac.kr

Importance of Nodes ² Question: which nodes are important among a large number of connected nodes? Centrality analysis can provides answers with measures that define the importance of nodes ² Different Centrality Analysis a) Centrality based on the degree information of a node Degree Centrality Eigenvector (or Spectral) Centrality b) Centrality based on the geodesic (i.e., shortest path) of nodes Closeness Centrality Betweenness Centrality 2

Degree Centrality ² Degree Centrality Importance of a node is determined by the number of nodes adjacent to it High-degree nodes naturally have more impact to reach a larger population than other nodes within the same network Degree Centrality Normalized Degree Centrality where n is the number of nodes in a network 3

Degree Centrality ² Degree Centrality Degree centrality of v 1 is 3 Normalized degree centrality of v 1 is 3 / (9-1) = 3/8 4

Degree Centrality ² Degree Centrality It fails to capture the centrality in the following graphs. In most cases, it fails to capture ability to broker between groups where information is originated from 5

Betweenness Centrality ² Betweenness Centrality It measures the number of shortest paths in a network that will pass a node. Nodes with high betweenness play a key role in the communication within the network Betweenness centrality of a node where C * v & =, σ "# v & σ "# -. /- 0 /- 1 3, "5# Ø σ "# is the total number of shortest paths between nodes v " and v # Ø σ "# (v & ) is the number of shortest paths between nodes v " and v # that pass along the node v &. 6

Betweenness Centrality ² Betweenness Centrality σ 19 = 2 ç 1-4-5-7-9 and 1-4-6-7-9 σ 19 (4) = 2, and σ 19 (5) = 1 C B (4) = 15 all shortest paths from {1, 2, 3} to {5, 6, 7, 8, 9} have to pass v 4 C B (5) = 6 All the shortest paths from node {1, 2, 3, 4} to nodes {7, 8, 9} have to pass either v 5 or v 6 Betweenness centrality of all nodes 7

Betweenness Centrality ² Betweenness Centrality Maximum value of C B (v i ) in an undirected network with n nodes 6 7 8 9 1 5 2 3 C B (v 5 ) = 8 * 7 / 2 = 28 4 s=1 s=2 s=3 s=4 s=6 s=7 s=8 t=2 1/1 t=3 1/1 1/1 t=4 1/1 1/1 1/1 t=6 1/1 1/1 1/1 1/1 t=7 1/1 1/1 1/1 1/1 1/1 t=8 1/1 1/1 1/1 1/1 1/1 1/1 t=9 1/1 1/1 1/1 1/1 1/1 1/1 1/1 Normalized betweenness centrality 8

Betweenness Centrality ² Betweenness Centrality Example 1 (non-normalized version) A B C D E A lies between no two other vertices B lies between A and 3 other vertices: C, D, and E C lies between 4 pairs of vertices (A,D),(A,E),(B,D),(B,E) note that there are no alternate paths for these pairs to take, so C gets full credit 9

Betweenness Centrality ² Betweenness Centrality Example 2 (non-normalized version) 10

Betweenness Centrality ² Betweenness Centrality Example 3 (non-normalized version) 11

Betweenness Centrality ² Betweenness Centrality Nodes are sized by degree, and colored by betweenness Can you spot nodes with high betweenness but relatively low degree? What about high degree but relatively 12 low betweenness?

Closeness Centrality ² What if it s not so important to have many direct friends? or be between others. But one still wants to be in the middle of things, not too far from the center ² Closeness Centrality It measures how close a node is to all the other nodes It describes the efficiency of information propagation from a node to all the others It involves the computation of the average distance of one node to all the other nodes Closeness Centrality where n is the number of nodes, and g(v i, v j ) denotes the geodesic distance between nodes v i and v j. 13

² Closeness Centrality Closeness Centrality Closeness centrality of v 3 and v 4 We conclude that v 4 is more central than v 3. 14

Closeness Centrality ² degree number of connections denoted by size ² closeness length of shortest path to all others denoted by color 15

Eigenvector Centrality ² Eigenvector Centrality A node s importance is defined by its adjacent nodes importance. Conceptually, Let x denote the eigenvector centrality from v 1 to v n. Then, the above equation can be written as in a matrix form Equivalently, we can write where λ is a constant It follows that Thus x is an eigenvector of the adjacency matrix A. 16

Eigenvector Centrality ² Eigenvector Centrality How to get the top eigenvector x of the adjacency matrix A? Transition matrix A7 = Column-Normalized Adjacency Matrix A = A7 = Transition matrix is constructed based on the adjacency matrix by normalizing each column to a sum of 1: An entry A7 ij denotes the probability of transition from v j to v i 17

Eigenvector Centrality ² Eigenvector Centrality How to get the top eigenvector x of the adjacency matrix A? It can be computed by the power method Ø i.e., repeatedly left-multiplying a non-negative vector x with A7 Ø Ø x will be converged A7 = 18

Eigenvector Centrality ² Algorithmic Complexity Degree centrality & Eigenvector centrality è Low Complexity Closeness centrality & Betweenness centrality è High Complexity For large-scale networks Efficient computation of centrality is critical and requires further research ² Centrality Summary 19

Importance of Nodes ² Centralization: how equal are the nodes? How much variation is there in the centrality scores among the nodes? Freeman s general formula for centralization C D = g C D (n * ) C D (i) i=1[ ] [(N 1)(N 2)] maximum value in the network C 8 : Degree Centralization C 9 : Closeness Centralization C * : Betweenness Centralization C : : Eigenvector Centralization 20

² Degree Centralization Examples Importance of Nodes C D = 0.167 C D = 1.0 C D = 0.167 21

² Degree Centralization Examples Importance of Nodes financial trading networks high centralization: one node trading with many others low centralization: trades are more evenly distributed 22

연구사례 ² 국내트위터이용자의관계분석에관한연구, 양동선, 한연희, 한국통신학회 2011 년도동계종합학술대회 용어정의 ( 개인을기준으로 ) 팔로워 (Follower) Ø 그개인을따르는사람 (=Twitter s followers) 프렌드 (Friend) Ø 그개인이따르는사람 (=Twitter s followings) 데이터수집 트위터 Search API 를사용하여다음과같은계정을지닌 9351 명의사용자정보를수집 Ø 이용자지역정보에 Korea 를포함, 지역명을한글로기재, Timezone 을 Seoul 로설정 (2011 년도 8 월기준 ) 위와같은사용자정보중팔로워 / 프렌드관계를수집할수없게보호된계정 274 명을제외한 9077 명을대상으로분석 23

연구사례 ² 국내트위터이용자의관계분석에관한연구, 양동선, 한연희, 한국통신학회 2011 년도동계종합학술대회 두그래프모두 Power-law Distribution 형태를보이고있으며, 이는국내트위터이용자의관계내에서도롱테일 (Long-tail) 현상이나타나고있음을의미한다. 24

연구사례 ² 국내트위터이용자의관계분석에관한연구, 양동선, 한연희, 한국통신학회 2011 년도동계종합학술대회 팔로워 / 프렌드관계의양이많아질수록상호팔로잉하는비율이높아짐을나타내고있다. 팔로워수와프렌드수사이에는양의상관관계가존재한다 25