Computer and Information Science; Vol. 9, No. 4; 2016 ISSN 1913-8989 E-ISSN 1913-8997

Published by Canadian Center of Science and Education

44

Design and Simulation Analysis of Multi-Agent Online Dissemination Model on the Basis of "The Spiral of Silence" Theory

Qiuyun Miao1 & Yang Shen1 1 School of Economics and Management, Nanjing University of Aeronautics and Astronautic, China Correspondence: Yang Shen, School of Economics and Management, Nanjing University of Aeronautics and Astronautic, Nanjing, China. Tel: 1-381-390-5870. E-mail: shen.y@nuaa.edu.cn Received: September 30, 2015 Accepted: October 16, 2015 Online Published: October 31, 2016 doi:10.5539/cis.v9n4p44 URL: http://dx.doi.org/10.5539/cis.v9n4p44 Abstract This article studies the classical theory of communications—the Spiral of Silence. The key factors are extracted for multi-agent modeling and simulation with Netlogo. It turns out that opinion climate is the external formation of the Spiral of Silence. Besides, the size of opinion climate has no direct effects on the Spiral of Silence. Keywords: the Spiral of Silence, multi-agent modeling, simulation, Netlogo 1. Introduction The innovation of "The Spiral of Silence" theory has attracted many researchers under today’s network communication environment, some scholars believe that the the Spiral of Silence in cyberspace is weakening which plays a new role in network media. GAO Xianchun et al. think that: "Mass media and specific individual communities make a difference in the new media environment constructed by Internet, mobile phones and others which are in the construction of specific event. And they both trigger the double helix of silence. In other words, mass media and individual communities form spirals respectively, meanwhile, both games lead to opinion cassette which is the factor affecting individual perception. As a result, the double helix of silence influences individual expressions of opinion. "(GAO Xianchun et al,2014) Yang Zhibiao supports the thory despite of its assumptions and limitations. In his opinion, it is still available in network environment. What is more, "the spiral of silence" is alive with "Against the Spiral of Silence". However, other scholars believe that the Spiral of Silence is still spinning in the new era. To explore the views of "Gay Bullying", Sherice Gearhart et al. investigate 760 participants. It was found that the the Spiral of Silence still exists in social networks, besides, personal ideology will also have an impact on the willingness of expression (Sherice Gearhart et al, pp. 18-36, 2014. ). Li Chuansheng points out that the Spiral of Silence still works because of the fear of isolation and group pressure in the network. What is more, the Internet has a character of anonymity, but net users’ behavior is not unrestricted and complete freedom does not exist which due to group norm and social norm in the network community. That is the reason why the Spiral of Silence is still in the spin (Li Chuansheng, pp. 4-12, 2014). Na Yeon Lee et al. carry out a series of studies on controversial topics among journalists who work for nine national newspapers and two networks broadcasting corporation through the empirical research. It turns out that the Spiral of Silence still exists, especially in the social networking site—Twitter (YL Na, pp. 443-458, 2015). Cheng Yao also points out that because of unconverted habits of netizens, squelch on the Internet and attention which the country begins paying to supervising public opinion online, the Spiral of Silence will coexist with public opinion on the network for a long time in a more subtle way. Moreover, according to a report released by the Pew Research Center and the Rutgers University, social media such as Twitter and Facebook will contain the diversification of views, hinder public affairs debates and limit people to state their views. Especially when they find that their opinions are different the fear of isolation; climate of opinion, and quasi-statistical sensewith friends’. The report also said that very few people who use social media regularly express different points of view offline. In order to study the impacts mechanisms of the Spiral of Silence, Scholars adopt empirical research through questionnaire to explore the impact of " the climate of opinion " – one factor of the Spiral of Silence. In Chang Ning’s research, it is found that post-90 college students adhere to the "expression"of the highest degree in real life. Unexpectedly, under the anonymous network, post-90 college students tend be more silence. The study also shows that the anonymity of the network is not the main factor to promote the expression. Zhou Baohua

cis.ccsenet.

investigateanalysis, dconsensus students othose of paAbove all,opinion inmechanismcharacterispropagatiosimulate th2. FramewThe Spiralhow peopexpressionsilence is wIn 1980, Ethought thfear of humsilence becmore willihypothese(1)Individu(2)The ind(3)The fea(4)The restheir views(5)This hypublic opinDue to thisolation, pcycle, one 3. Modelin3.1ModeliAccordingopinion an

org

es the willingndifferent clima

with the majon the Internetarents’, the net, the Spiral of

n the new era,m of the Spirastics of the Spon model of thhe model. Finawork of the Dl of Silence thele decide whe

n is associated widely used inElisabeth Noe

he spiral of sileman beings. Wcause of fear oing to expresss in summary. ual fear of isol

dividual evaluaar of isolation isults of evaluas. ypothesis relatnion and so on

he fear of isolpeople will tryvoice is grow

ng and Simulng based on th

g to Neumannnd quasi-statist

ness of collegete of opinion hority in socie. However, optwork has a sigf Silence may , but it still wal of Silence bpiral of Silencehe Spiral of Sially, the modeligest Format eory put forwaether or not twith external

n media and coelle-Neumann ence comes fr

When people foof isolation. Ons himself/herse

lation; ates the climateis used to deal ation affect ind

tes to four hypn. (Noelle-Neuation, the agey to seek the ming, while the ation based ohe Spiral of Siln’s theory, thertical sense. The

Fi

Computer an

e students and has different ety and those o

pinions of collgnificant positinot be fully c

works. At the by multi-agente from relativeilence by usinl will be valida

ard by Germanto express thepressure, such

ommunication rpublished Th

rom the natureound himself/hn the contrary,elf. In this bo

e of opinion cowith individua

dividual behav

potheses aboveumann, pp. 143ents will chanmainstream viother tends to n the Theorylence re are three me paper summa

igure 1. Model

nd Information S

45

its factors by effects on colleonline has no lege students aive effect on c

consistent withsame time, tht modeling. Tely objective p

ng multi-agent ated.

n communicatieir own true fh as the mediaresearch to des

he Spiral of Se of the societyherself is a me when a person

ook, she thoug

onstantly becauals who deviat

vior and opinio

e and they com3-158, 1977)nge their opiniiews and obsebe silence wh

main factors oarizes model a

l of the Spiral

Science

sampling survege students w

obvious effecare not the sam

college studenth the characterhere are few pThe paper attemperspectives. Amodeling. Th

on scholar Elifeelings. Accoa and the viewscribe the silen

Silence: Publicy, at the sameember of the mn is a member

ght that the the

use of fear of ite from the maon in public, e

mbine to expl

ions accordingrve the rally o

hich is called th

f the theory: as shown in Fig

of Silence.

vey. Based on which is suggescts on the expme as those ots to express thristics of dissepapers explorinmpts to descriAt first, this phen, the paper

sabeth Noelle-ording to the ts of people aronce of controvc Opinion-Oue time, it comeminority, he/shr of the majoriteory was mad

isolation. ainstream by thespecially by r

ain the format

g to majority of all kinds ofhe Spiral of Sil

the fear of isgure1.

Vol. 9, No. 4;

the test resultssted. The degrpression of coof main mediuheir opinions. emination of png the propagibe communic

paper will consadopts Netlog

-Neumann exptheory, the wiound. The spir

versial events. ur Social Skines from entrenhe would tend ty, he/she woude up of five

he society. revealing or h

tion and chang

opinion. To af views. Repealence.

olation; clima

2016

s and ee of

ollege um or

ublic ation ating struct go to

plains ill of ral of , she

nched to be

uld be basic

iding

ge of

avoid at the

ate of

cis.ccsenet.

Accordingand "consithis paper Hypothesipositive opratio of poHypothesian overwhopinion bymechanismHypothesiThere are t

The agentchanges, inincludes thin the prog"radius". Pcertain "p"count-aro3.2 Simula(1) PropertThe fear osystem in tof Spiral ochanges ininfluencesproperties Table 1. In

AgenTurtl

Patch

org

g to the model,istency with thraises some brs 1: There arepinion and neg

ositive and negs 2: There is o

helming view iy which the mm of the Spirals 3: In given ctwo main agen

t "turtle" is mndividual afraihe followings: gram) and "geProperties are patch" in ordound-num" is nation based on ties Settings of isolation anthe theory whiof silence onlinn this network. each other thrin Table I:

ntroduction of nt variableles Initial-pe

Initial-ch

r-afraid-ig-afraid-

hes red-turtlegreen-turradius

, the ultimate she majority" crief and necesse only two sidgative opinion

gative commenonly one kind in the dissemin

majority opiniol of Silence weconditions, Indnts in the const

made up of twoid-index will c"change-afrai

et-interest". Prused to counter to simulat

needed to achiethe Spiral of S

nd external clich prompts thne exists in co Attitude chanrough interacti

Property Settine name eople

hance

index -index es rtles

Computer an

source of the Scan affect the wsary assumptiodes in the dissn. At the saments.

of the majoritnation of publion can be randell. dividual views truction of sim

Figure 2

o attributes: achange accordid-index"、"ch

roperties of thet the number ote the climateeve the goal.Silence

limate of opinhe views of somomplex social nges of netizenions which cau

ngs explanationvalue scope representing value scope representing representing climate chanrepresenting "radius" arouvalue scope representing can be affect

nd Information S

46

Spiral of Silenwilling of expons: semination of

e time, climate

ty opinion onliic opinion. Thudomly decided

may change almulation model

2. Class Diagra

afraid-index aningly and agenhange-color" (Ce agent "patchof green and ree of opinion

nion affect theme events, evanetwork. The

ns can be reflecuses individual

[0,1000] the number of

[0,1] the probabilityred or green t

nges. the number o

und a certain "[1,10] the radius of

ted

Science

nce is fear of ispression directl

f public opinioe of opinion ca

ine. On the baus, adjusting td in the Intern

long with the sl, as shown in F

am

nd interest. Wnts may changColor changesh" consist of "ed turtles with

around a ce

e circumstancaluations and o

"Agent" can bcted by the chal to change op

f people in the

y distribution turtles' fear ind

of green and repatch"

the climate of

solation, but "cly. In order to

on. That is to an be simulate

asis of the theothe proportion net can simula

spread of publFigure 2.

When external e their opinion

s indicate the c"green-turtles"hin a radius ofertain "turtle"

es of informaopinions in socbe seen as maange of the "Aginions. The stu

e initial state

of initial state dex and it will

ed turtles with

f opinion in w

Vol. 9, No. 4;

climate of opinfacilitate rese

say there are ed by adjustin

ory, there is alof two sides o

ate the propag

ic opinion.

climate of opns. Setting metchanging of op、"red-turtles"f "radius" arou" and the me

ation disseminciety. The evolin element of gent". The "Agudy designs ag

l change as the

hin a radius of

which the turtle

2016

nion" earch,

only g the

ways of the ation

inion thods inion " and und a ethod

ation ution view gent"

gent’s

e

f

e

cis.ccsenet.org Computer and Information Science Vol. 9, No. 4; 2016

47

Most Internet users will try to avoid isolation caused by separate attitudes and views. When agents find their own views are different from those of most Internet users or those of the mainstream media, they will not tend to express their opinions. Instead, they may choose to change views to correspond to the mainstream views online. (2) Settings of initial state The slider--initial-chance is set to represent the distribution of initial state. The proportion of positive and negative opinions will vary when setting different ratios. Besides, The "initial-people" stands for the number of people who know the event and express their views. The "radius" is on behalf of the size of the climate of opinion in which the agent can be affected. According to the spiral of silence, the fear of isolation is inherent nature of the people, thus the initial values of "r-afraid-index" and "g-afraid-index" are set to 50 each. (3) Settings of interaction rules a. Collecting Statistics of climate of opinion around some individual: through Netlogo program, the climate of opinion can be collected automatically. (Here’s the code.)

to count-around-num ask turtles [set red-turtles(sum[count turtles with [color = red] ]of turtles in-radius radius) set green-turtles( sum[count turtles with [color = green] ]of turtles in-radius radius)]

end b. Changing Attitudes: when the value of the fear of isolation reaches a certain critical value, Internet users may change their views. What is more, when corresponding to the majority, agents will be willing to express their opinions and try to influence other agents. That’s why there is a process called "hatch", here’s the code.

to change-color if red-turtles > green-turtles [ if color = green and random r-afraid-index < 35 [set color red fd random 10 hatch 1 [set color red fd random 10 set r-afraid-index 50 ]] end

c. The ultimate source--"fear of isolation": the fear index will change accordingly when the individual is affected by the surroundings. In addition, different agents have different degrees of changes on the fear index. This part of the code is as follows:

to change-afraid-index if red-turtles > green-turtles [set r-afraid-index(r-afraid-index - random 10) set g-afraid-index(g-afraid-index + random 10)] if red-turtles < green-turtles [set g-afraid-index(g-afraid-index - random 10) set r-afraid-index(r-afraid-index + random 10)] end

3.3 Analysis of Experimental Results The initial value of " initial-people" is set to 880 which represents the number of people knowing the event at the first time. The initial value of " radius" is set to 1. The percentage of two sides is randomly decided and different percentages have direct impacts on future climate trends. In this paper, the values of initial-chance are 0.1, 0.3, 0.5, 0.7 and 0.9 respectively. Parameters are shown in Table II.. Table 2. Parameter 1

No. initial-people initial-chance radius r-afraid-index g-afraid-index 1 880 0.1 1 50 50 2 880 0.3 1 50 50 3 880 0.5 1 50 50 4 880 0.7 1 50 50 5 880 0.9 1 50 50

cis.ccsenet.

Then, Presto change,follows.

org

Experim

ss the "go" but, and the corre

Experim

ment 1

ExperimFigure 3. initia

tton to start ruesponding grap

ment 1

Computer an

Ex

ment 3 al interfaces co

unning the progphs also began

E

nd Information S

48

xperiment 2

Eorresponding t

gram. Agents n to be painte

Experiment 2

Science

Experiment 4to different pro

with different ed to show the

Experiment

oportions

views in the ie change of di

Experim

Vol. 9, No. 4;

3

interface will bifferent opinion

ment 3

2016

begin ns as

cis.ccsenet.

As can be ratios in thopinion resilence dothat is to spicture abominority oIn order tothe climatethe study, Table 3. Pa

Through th

org

seen in Figurehe initial stateepresents signes not work. I

say, the distribuove that the cl

opinion weakeno study the rele of opinion inas shown belo

arameter 2 No. ini1 882 883 884 88

hese experime

Expe

Experim

e 4., characterie (0.1, 0.3, 0.7ificantly fluctIt can be conclution of viewslimate of opinin. lationship betwn which the tu

ow.

itial-people 0 0 0 0 nts, correspon

eriment 1

Computer an

ment 4 Figure 4. c

istics of spiral 7 or 0.9): Thetuate decliningluded that unes affects the foion is able to m

ween the theorurtle can be aff

initial-chance0.1 0.1 0.1 0.1

nding changes o

nd Information S

49

change curve c

of silence are e “majority” og. When the veven distributioormation of spimake the majo

ry and the parfected, five gro

e radius r-1 54 57 510 5

of curves are s

Science

Experimenchart

reflected clearpinion keeps value of “initon of opinionsiral of silence.ority opinion s

rameter—radiuoups of experi

-afraid-index0 0 0 0

shown as below

Experimen

nt 5

rly regardless oincreasing, whial-chance” iss is necessary . Besides, it castrengthen, wh

us which standimental param

g-afraid-in50 50 50 50

w:

nt 2

Vol. 9, No. 4;

of different ophile the “minos 0.5, the spirfor the theory

an be seen fromhile it can mak

ds for the radimeters are set u

ndex

2016

inion ority” al of

y, and m the ke the

us of up for

cis.ccsenet.

It is easy tthe inner ma few comresults are Table 4. R

R(1) represrepresents represents There is noof climate 4. ConclusThe paper analyses itopinion grexternal faspiral of siReferenceChang, N.

of CoCheng, Y.

403-4Gao, X. C

mediaLi, C. S. (2Na, Y. L.

relatio

org

to find that themechanism andmparative tests

summarized i

Research betwe

sents the numbthe number ored-turtles' av

o direct link bof opinion hassion is based on th

t. We can drawrow in strengtactor of the spiilence. es (2015). Comp

ommunication (2013). Study

405. C. et al. (2014)a environment2014). The app(2015). Twee

onship with th

Experiment 3Figure 5. c

e size of climad the relationsh were conductin Table IV :

een radius and radius 1 4 7 10

ber of red-turtf green-turtles

verage rate of r

etween the rads no direct imp

he framework ow the followingth while the miral of silence;

plex PerformanUniversity of Cying the ‘spira

. From "passivt. Press, 9, 43-plicability of speting the publhe public. Infor

Computer an

3 change curves

ate of opinion hip between thted. Apart from

opinions chanR(1) R(2) 792 2014792 1998792 1975792 1987

tles at first; R(

s at first; G(2) rrise; G(AVG) rep

dius and the chpact on the form

of the spiral ofg conclusions

minority opinio(3) The size o

nce of the SpirChina, 6, 109-al of silence’

ve silence" to 54. piral of silencelic: journalistsrmation Comm

nd Information S

50

corresponding

does not affeche size of climm the value o

nge rate G(1) G(1)

88 69 88 70 88 75 88 71

(2) represents tepresents the n

presents green-

hange rate of omation of spir

f silence. At thexperiments: (

on tend to die of climate of op

ral of silence t113. status seen fro

"active interac

e in the networ' Twitter use,

munication & S

Science

Experg to different ra

ct the formatiomate of opinionof “radius” oth

R(AVG) G(AV

174.57 2.71172.28 2.57169 1.86170.71 2.42

the number ofnumber of gre-turtles' averag

opinion. Thus,ral of silence.

e same time, th(1) The spiral out finally; (2

pinion has no

theory in the n

om BBS netw

ction": "doubl

rk era. Spread attitudes tow

Society, 18(4),

riment 4 adius

n of spiral of n and the rate oher parameters

VG)

1 7 6 2 f red-turtles afeen-turtles aftege rate of rise.

it can be conc

he study simulof silence can2) The climatedirect effects o

ew media env

work environm

e spiral of sile

and Copyrighward the public

443-458.

Vol. 9, No. 4;

silence. To expof opinion chan are invariant.

fter 7 secondser 7 seconds; R

cluded that the

lates the theoryn make the maje of opinion ion the formatio

ironment”, Jou

ment. Jin Tian,

ence" effect in

ht, 8, 4-12. c's tweets, and

2016

plore nges, . The

; G(1)

R(AVG)

e size

y and jority is the on of

urnal

316,

n new

d the

cis.ccsenet.org Computer and Information Science Vol. 9, No. 4; 2016

51

Noelle-Neumann. (1977). Turbulences in the Climate of Opinion: Methodological Applications of the spiral of silence theory. Public Opinion Quarterly, 41(2), 143-158.

Sherice, G. et al. (2014). Gay Bullying and Online Opinion Expression: Testing Spiral of Silence in the Social Media Environment. Social Science Computer Review, 32(1), 18-36.

Yang, Z. B. (2014). New ‘Silence Spiral’ Theory in the Environment of Internet Communication. Journal of Harbin University, 35(7), 95-98.

Zhou, B. H. (2014). Empirical study of college students ' network expression and factors—the ‘Spiral of silence’ and the ‘climate of opinion’ as the core. Contemporary Communication, 5, 34-37.

Copyrights Copyright for this article is retained by the author(s), with first publication rights granted to the journal.

This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/).