2024年3月16日发(作者：丰田锐志2016款)

Ef?cientAlgorithmsforCitationNetworkAnalysis

VladimirBatagelj

UniversityofLjubljana,DepartmentofMathematics,

Jadranska19,1111Ljubljana,Slovenia

e-mail:lj@

Abstract

Inthepaperveryef?cient,linearinnumberofarcs,algorithmsfordeterminingHum-

monandDoreian’sarcweightsSPLCandSPNPincitationnetworkareproposed,and

acyclicityproblemin

oachtoidentifyonthebasisofarcweightsanim-

portantsmallsubnetworkisproposedandillustratedonthecitationnetworksofSOM(self

organizingmaps)literatureandUSpatents.

Keywords:largenetwork,acyclic,citationnetwork,mainpath,CPMpath,arcweight,

algorithm,selforganizingmaps,patent

a

r

X

i

v

:

c

s

/

0

3

0

9

0

2

3

v

1

[

c

s

.

D

L

]

1

4

S

e

p

2

0

0

3

1Introduction

ThecitationnetworkanalysisstartedwiththepaperofGar?eldetal.(1964)[10]inwhichthe

inpaper,onthe

exampleofAsimov’shistoryofDNA[1],itwasshownthattheanalysis”demonstratedahigh

degreeofcoincidencebetweenanhistorian’saccountofeventsandthecitationalrelationship

betweentheseevents”.Anearlyoverviewofpossibleapplicationsofgraphtheoryincitation

networkanalysiswasmadein1965byGarner[13].

ThenextimportantstepwasmadebyHummonandDoreian(1989)[14,15,16].They

proposedthreeindices(NPPC,SPLC,SPNP)–weightsofarcsthatprovideuswithautomatic

waytoidentifythe(most)importantpartofthecitationnetwork–themainpathanalysis.

howtoef?cientlycomputetheHummon

andDoreian’sweights,sothattheycanbeusedalsoforanalysisofverylargecitationnetworks

sthissometheoreticalpropertiesoftheHummonand

Doreian’sweightsarepresented.

TheproposedmethodsareimplementedinPajek–aprogram,forWindows(32bit),for

eelyavailable,fornoncommercialuse,atitshomepage[4].

ForbasicnotionsofgraphtheoryseeWilsonandWatkins[18].

1

Table1:Citationnetworkcharacteristics

network

DNA

223

Smallworld

1059

Cocitation

3084

Kroto

4470

Zewail

8843

Desalination

3774768

25751

1

54253

7

1411

3764117

31950

2

101

8212

115

32

4929

1

0

3704

5

63

73

770

1988

1

35

2678

1

24

166

1098

12

0

m

60

1

163

1024

1

32

166

735

38

4

1

n

0

1

218

1

28

90

105

6

0

2

k

C

3

16

60

232

2

2

0

?

in

7

134

0

0

0

2

0

0

0

4

0

更多推荐

锐志,作者