Buzzer Works on Human-Machine Agency Role

in Securing Indonesia's Voters Precinct

Henry Sianipar

, Mediana Handayani

and Ressa Uli Patrissia

Department of Communication Science, Prof. DR. Moestopo Beragama University, Hang Lekir, Jakarta, Indonesia

Keywords: Buzzer; Political Algorithm; Polarization; Trending Topic, Human-Machine Agency.

Abstract: This paper investigates political polarization observed on Twitter during the 2019 Indonesia presidential

election to illuminate a complex and organized human-machine interaction driven by the personalized

algorithm and its impact in securing designated voters. In this research, the theory of Interactive Media Effects

(TIME) leverages to verify that personalized algorithm provides a myriad opportunity for buzzers to

commercially politicize using the bot as machine agency in targeting undecided and swing voters for a

Trending Topic Operation. An autonomous application NodeXL and Twitter API successfully acquired

328.474 tweets as the source for social network analysis. Snowball technique in obtaining data interviews

from 6 field experts served as triangulation along with related literature support. A study on the 2019

Indonesia Presidential election revealed that buzzer activity played a significant role in securing votes by

deceiving the Twitter algorithm. The human agency has proven more impactful than machine agency in

creating trending topics. Findings also show that bot use was still widespread among 2019 Indonesia

presidential candidacies. These findings redefine the electoral voter's sphere in Indonesia as the new

promising precinct emerged and raise further concerns of the possibility of instigating a specific behavior

through manipulating one's political preference.

1 INTRODUCTION

On April 17, 2019, 192 million Indonesians who were

eligible to vote in a general election chose the nation's

destiny. In an essential democratic exercise,

presidential, parliamentary, and regional elections are

conducted simultaneously on the same day for the

first time in history. Current President Joko Widodo

is running against former military commander

Prabowo Subianto in this election.

The Elections Commission reports that

approximately 80 million individuals, or 40% of

eligible voters, are between 17 and 35. Because

young people are the most crucial voters, social media

is the best method to reach them. In Indonesia, social

media reigns supreme (the nation has the world's

fourth-largest Facebook user base). Since the

election, the country has been plagued by fake news

problems, with political false news and

disinformation rising by 61 percent between

https://orcid.org/0000-0003-2791-3523

https://orcid.org/0000-0001-6270-0158

https://orcid.org/0000-0003-1480-6753

December 2018 and January 2019. Despite

Facebook's dominance, Twitter has been extremely

popular among social media users throughout

election season owing to its ease of sending political

instructions and reachability via trending topics.

Political participation, for example, has changed as a

consequence of social media. Because of changes in

the platform for political mobilization, the emergence

of uneven distribution of media power, and a flood of

information on a mobile phone, some campaigns

utilize crowd mobilization as an offline tactic, while

others use a virtual environment platform.

(Akmaliah, 2018). The massive number of social

media users, coupled with a lack of digital literacy,

further obscures the purpose of social media; in

reality, many have a negative effect owing to the

abundance of false and untrustworthy news.

Political interests and propaganda of

contemporary political objectives are often expressed

via social media through sponsored tweets,

352

Sianipar, H., Handayani, M. and Uli Patrissia, R.

Buzzer Works on Human-Machine Agency Role in Securing Indonesia’s Voters Precinct.

DOI: 10.5220/0010921600003255

In Proceedings of the 3rd International Conference on Applied Economics and Social Science (ICAESS 2021), pages 352-359

ISBN: 978-989-758-605-7

sometimes known as the buzzer. When Twitter burst

in popularity in 2006, the term "Buzzer" became well-

known. Along the process, Twitter coined new terms

such as netizens, followers, influencers, twitwar, and

buzzers, which became popular in 2009 for brand

promotion. Since then, social media usage in

Indonesia has skyrocketed, as has the flow of

information in the media (Dimedjo, 2019). Twitwar

especially (means of combating issues through social

media) driven by the buzzer, on the other hand, has

begun to emerge, polarizing, and dividing society by

bringing up the subject of religious and racial feeling

(SARA), which was previously deemed taboo in

political contestation in Indonesia.

During the 2019 election season, there were two

types of buzzers: volunteer or unpaid buzzers and

paid buzzers who received specific requests.

(Saraswati, 2018). According to research was done by

the Center for Innovation Policy and Governance

(Camil et al., 2017), the buzzer generates in two ways,

naturally and organically driven by market demand.

There are two main reasons why a specific person or

account is buzzing. The flow of money defines

commercial incentives, while ideology or a feeling of

fulfillment defines volunteer incentives.

The presence of a political buzzer has divided the

two sides, with supporters of Jokowi called "cebong"

and opponents branded "kampret." The buzzer

highlights the problem of insults, hate, and

provocation aimed at one another by the two groups,

who are progressively separating themselves from the

intended democratic atmosphere.

A method for trending topics is part of Twitter's

algorithm. The buzzer works by distributing news and

tricking trending algorithms, then filtering the content

that shows on a user's account page depending on

views and likes. The number of posts and re-posted

(retweeted) in a given time by netizens from various

geographic areas, and a minimum of about a thousand

tweets at the same time within the scattered area

tagging are some of the criteria by which one can

create trending topics at the national level in a short

period. Furthermore, network functions such as

muting, unfollowing, and unfriending are linked to

polarization. Only those who have the same

viewpoint will be addressed. Because of this, both

kampret and cebong arise.

Another research conducted by Samantha

Bradshaw and Philip N Howard (Bradshaw, n.d.)

revealed that buzzers are utilized in Indonesia for

engagement behaviors, tools, and resources for

propaganda distribution. One of the ways for negating

the message you want to magnify is to use a bot. The

buzzer will communicate bot account tales and use

automated accounts. Bot accounts will create posts

from unknown sources. Once information is made

public, it gets picked up by influencers and buzzers.

We discovered that a study on buzzer activity in

the Indonesian presidential election restricted the

reach of social media and its polarizing effect. Based

on this context, we examine how political commercial

and volunteer buzzers work in Twitter and the use of

human and machine agency interaction to secure

voter choices.

2 LITERATURE REVIEW

2.1 Online Disinhibition Effect on

Political Interest

The concept of a new public sphere, popularized by

Jurgen Habemas' book The Structural Transformation

of the Public Sphere: An Inquiry into a Category of

Bourgeois Society (Habermas, 1989), is essential to

begin the literature review of this research because it

explains the concept of space created by a collection

of people. Specific individuals (private individuals)

seem to have been formed as a kind of attitude against

governmental authority. A public place free of ruler

dominance appears to be worthy of being

incorporated in an internet that is readily available

and free of state and commercial constraints, allowing

people to engage in political discussions. The low

degree of political literacy and reading culture in

Indonesian society adds to the negative aspects of

social media. It is elementary to spread false news or

hoaxes. Hoax news spreads fast via the fingertips of

regular people who don't know what information he

just delivered by exploiting the innocence of ordinary

people. This is due to people may communicate

something in the realm of media that is socially

impossible in real life. It also demonstrates the user's

lack of self-control on social networking, known as

the online disinhibition effect: the ability to freely

communicate anything you want to say (Suler, 2017).

The use of social media to disseminate information to

the public or voters in elections is seen as an effective

and crucial step, particularly in shaping political

opinion and agenda-setting (Woolley & Howard,

2018). However, though social media has a

significant impact on voter political engagement, it

impacts political knowledge (Dimitrova et al., 2014).

The presence of social media in political campaigns

in Indonesia has altered the tactics and strategy of

winning both 2019 presidential and vice-presidential

candidates, as shown by the development of a

dedicated team that handles social media. The use of

social media as a channel for community members to

Buzzer Works on Human-Machine Agency Role in Securing Indonesia’s Voters Precinct

353

communicate during social and political protests can

potentially upset the status quo by changing policies

or causing structural changes. The Arab Spring in

2011 was an example of a regime's demise due to

social media encouragement (Tufekci, 2017) by

allowing for public involvement, which social media

aids movements are calling for political change (Lim,

2012). As a media tool, Twitter disseminates general

information acquired by the community and most

impact political change via opinion formation

through a hash mark (#hashtag), which canonicalize

the subject, concentrate on themes, and assist

internet-based search engines (Syahputra, 2017).

From here, the dominance of online political

discourse on social media, more or less influenced by

the resources owned by each political party. Thus, it

is not surprising that political figures from major

political parties often dominate political discussions

online or on social media (Klinger, 2013).

2.2 Algorithm and Polarization

In general, three aspects become the main focus of

research in the parties/candidate's category: the

characteristics of political parties or candidates who

use Twitter as a political medium, how they use

Twitter as a political medium, and the effectiveness

of using Twitter as a political communication tool by

them (Jungherr, 2016). Moreover, Twitter is used by

politicians to produces a high buzz effect that the

mainstream media can amplify. In this case, tweets

are more of a reactive action and not a tool to predict

the contest (Murthy et al., 2015). Twitter becomes a

location that is regarded as the most appropriate for

gossiping even in cyberspace due to the social aspects

of people who want to gather and talk, debating

rumors or problems (Pohjonen & Udapa, 2017). The

gossip arena in cyberspace sometimes turns into war.

For instance, Twitter accounts affiliated with

presidential candidates such as @GarudaPrabowo,

@Gerindra, @Jokowi4me, @PDI_Perjuangan,

@Relawan_Jokowi, @FansGerindra, and other

accounts are often seen debating online (twitwar).

Not to mention communally managed accounts like

@PartaiSosmed, @99army, @Triomacan2000, who

are part of the campaign interests of the presidential

candidates, are often involved in twitwar defending

their presidential candidate. Hashtags are in close

relation to an algorithm. In simple terms, the

algorithm works in two stages: (1) knowing tastes and

preferences of social media account owners which

data can be found from various activities on social

media such as clicking, searching, or share social

media content to provide content according to your

sense of humor, liking, phobias, and even sexual

tendencies, and (2) the algorithm engine works by

classifying people who have in common: tastes,

ideologies, phobias, and so on.

Simply put, like-minded ideologies and tastes are

put together. The existence of traditional propaganda

containing lies and misinformation spread online is

algorithms (Woolley & Howard, 2018). Further, lack

of control and ambiguity about algorithmic

assessment may create algorithmic anxiety as

individuals are labeled and categorized by the

machine (Jhaver et al., 2018). Another negative

consequence of an algorithm is the Filter Bubble that

separates a person through personalization which

may undermine the internet's initial function as an

open platform for exchanging ideas, leaving us all in

an isolated, echoing world. The filter bubble also

creates a false consensus effect; that is, a person tends

to claim that others agree with him, and conclude his

opinion is the majority's conclusion (Pariser, 2011).

The harmful effects of filter bubbles are getting worse

due to the bad habits of netizens and media. The

media likes to make bombastic titles (clickbait), and

netizen's practices to share content without thorough

reading played a role in the enormous effect of this

bubble.

2.3 Buzzer Activity

Buzzer activity in Indonesia began to be used in 2009

for promotional interests. Buzzer involvement in

political events started used in 2012 during the

political contestation of the DKI Jakarta Pilkada. On

2014 presidential election, the use of buzzers is

increasingly being used in the political arena 2014

Presidential Election. Buzzer involvement in political

campaigns has contributed negatively to the image

and the meaning of society towards the buzzer.

Buzzers produced negative content and even hoaxes

on social media (Camil et al., 2017). Even after the

Reformation Order, the internet is still seen as one of

the media over which the government has little

influence in terms of social and political problems

(Setianto, 2015).

2.3.1 Types of Buzzers

Based on the motive (Camil et al., 2017), buzzer can

be divided into three, namely: (1) professional

commercial buzzer, which indicated by the flow of

funds, motives are purely for money and has nothing

to do with ideological or personal principles, also

have good command technical and reading skills

ICAESS 2021 - The International Conference on Applied Economics and Social Science

354

well, (2) voluntary ideological buzzer or neutral

buzzer works on an equal of ideological, political

views and goals to create a better situation for himself

and society, (3) paid ideological buzzer, which

support for one of the presidential candidates by

focusing more on winning the presidential candidate

where there is monetary reward for being obtained.

2.3.2 Human-Machine Agency

The nature of human communication alters as they

utilize media. Communication takes on new forms,

speeds, processes, scales, and even content due to

media. Every kind of media has the potential to

influence the user. Mass media is more than simply a

medium for gathering and disseminating information

that has the potential to affect viewers' ideas and

behaviors. It is not enough to study the contents of the

media to assess their effect; it is also necessary to

examine themselves and the character of the media

itself (McLuhan, n.d.). (Shyam Sundar, 2008)

discovered that the present machine agency plays a

more prominent role as an agent in determining the

message, which is now more decided by an algorithm.

Sundar's research has resulted in the proposal of a

dual framework, a synergy between human and

machine agency, with the application of Theory of

Interactivity Media Effects (TIME) to investigate

symbolically and allows the ability of media effects

TIME is based on four interactive media models,

namely: (a) The Interactivity Effects Model, (b)

Agency Model of Customization, (c) Motivational

Technology Model, and (d) Modality-Agency-

Interactivity-Navigability (MAIN), which set the

TIME signal route. Together these four models serve

to explain how interactive media technologies shape

perceptions of online action.

The term interactive in the context of media refers

to to users who can intervene directly and can change

the images and text they access. In new media, the

audience becomes a user compared to a viewer

(audience) or becomes a reader. TIME argues that the

nature of user engagement depends on the mediators

involved in specific interactions (Sundar et al., 2015).

Buzzer's action plays the human agency's role by

using bots (machine agency) during message

amplification. Bots (machine agencies) are used as

part of an amplification strategy towards trending

topics. Humans create bots. Bots are also designed to

interact with each other as if functioning as a regular

Twitter account. Commercial buzzer teams deploy

programmed bots (robots) to add tweets per

minute/hour and are expected to keep going up. Bots

Table 1: Four Interactive Media Models (The Handbook of

the Psychology of Communication Technology, 2015).

The

Interactiv

ity

Effects

Model

Agency

Model of

Customizati

Motivational

Technology

Model

Modality-

Agency-

Interactivi

ty-

Navigabili

(MAIN)

Source

interaction

leads to

more user

engageme

significant

(contributi

on) with

the media

increasing

user

capabilitie

customize,

curate, and

create

conten

Contingency-

enhancing

interactions,

modalities,

and

navigation

capabilities

messages,

self-

acknowledgm

ent, and

individual

exploration

that serve to

increase self-

recognition as

a news source

Navigation,

interaction,

and

customization

capabilities on

greater

intrinsic

motivation by

increasing user

competence,

interconnected

ness

Interfaces

modality,

agency,

interaction,

and

navigation

capabilities

that

makeup

perception

and user

experience

by using

cognitive

heuristics

about

content

quality and

credibility

that have followers and follow each other seem to be

involved in a natural chat/conversation so that the

conversation is recorded continuously, rotating

mutual interaction between them until the number of

discussions up and can become a trending topic.

Professional buzzer teams take advantage of trending

topics to assign hashtags #, attack/tweet

simultaneously at almost the same time, and playing

with public emotions to gain viral tweets.

3 RESEARCH METHOD

To achieve the purpose of this study, we conducted

social network analysis using autonomous

application NodeXL and Twitter API acquiring

328.474 tweets for analysis. This study employed a

qualitative descriptive approach, which was preceded

by exploratory research. To complete literature study,

we deploy snowball technique, a technique for

locating, selecting, and sampling in a continuous

network or chain of relationships. To collect data

from informants, we begin by conducting interviews

with informants we have been appointed with. We

requested a recommendation from the first informant,

whoever informant was available for the consultation.

In this instance, we will conduct in-depth and

structured interviews with the Presidential Candidate

01 Jokowi-social Amin's media team. Structured

Buzzer Works on Human-Machine Agency Role in Securing Indonesia’s Voters Precinct

355

interviews were conducted with key informants (team

controllers/coordinators), key informants (the

commander of the team that runs the bot), and the

buzzer team located throughout western Indonesia,

central Indonesia, and parts of eastern Indonesia.

Additionally, in-depth interviews were performed

with key informants of the presidential candidate 02

Prabowo-social Sandi's media team coordinator. This

is done in order to obtain information on the balance.

Additionally, informants or resource persons

interviewed originate from social media observers

and activists and prominent data activists who serve

as opinion-makers. Additionally, data triangulation

interviews with observers / digital communication

professionals will be done to ascertain data

confirmation and dependability.

4 RESULTS AND DISCUSSION

4.1 The Buzzer's Existence

From Twitter data processing, researchers gathered

328.474 tweets as the source for Social Network

Analysis. Researchers also interviewed key

informants such as political buzzer coordinators and

experts in media to do triangulation. From here, it is

known that most of the buzzers are recruited from

grassroots voters who bear the exact preference of

choice in the 2019 Presidential Election contestation.

Organic volunteers are recruited through

introductions between friends in the same profession

and organization. The buzzers were recruiting in

stages to fulfill one strategy: establishing a 7500

Twitter account from Aceh to Gorontalo. From here,

these 7500 accounts will be targeting to set 1000

accounts for each account (downline), projected to

secure in total 7,5 million voters (including undecided

and swing voters). The role pattern of these buzzers

aims to create a smooth transactional information

form up to the bottom simply. A facility such as a

salary and an internet quota are provided. Its mission

is to establish a dependable image and deliver

positive messages to the candidates. Aside from the

buzzer team that played a role on the positive side, it

is also known that another one played on the dark

side. This team specially recruited specialized

information technology skills to control a bot

(machine) and read through an algorithm. However,

some buzzers recruited voluntarily, consisting of

professionals in media and digital technology put

together under one organization. This organization

aimed to target voters' precinct weakness of one

candidate and volunteered to provide voice support

for social media operations, especially Twitter.

Militancy is not only built from official consolidation

in the campaign team but also Indonesia-wide buzzer

network. The result of triangulation shown that all

teams have recruited cyber troops and bot. The

number of cyber troops and bots from Presidential

Candidate 01 Jokowi is more than Presidential

Candidate 02 Prabowo. Moreover, the pattern of

cyber troops and bots for Presidential Candidate 01 is

more spread and fluid in various topics, while the bot

and cyber pattern of Presidential Candidate 02 is

more centralized and militant.

4.2 Trending Topic Operation

Buzzer mechanism needs to work and succeed at

Operation Trending Topic (OTT). This operation uses

the hashtag to categorize messages. If a hashtag

fulfills specific requirements, it may be featured in the

most popular trends (trending topic) list in the Twitter

timeline. One of the indications that a hashtag is

trending is a rise in the number of tweets over a

specific time period rather than the overall number of

tweets using the hashtag. Trending topic algorithm

based on how many hashtags were mentioned in

tweets in a short period of time. For example, a

hashtag reached the hot topic list because it was

referenced in messages (including retweets) in 3K

tweets within an hour. However, during the following

hour, that hashtag was used 3000 times in tweets.

Despite the large quantity, Twitter considers it to be a

constant. On the other hand, if a new hashtag emerges

and is referenced in a thousand tweets within an hour,

that hashtag will become trending. Placing a hot topic

on Twitter is a technique for winning over the masses'

psychology in the 2019 Pilpres contest. As it displays

which hashtags are being spoken about by tweeps,

Twitter's Trending Topic column is often used as a

battlefield for political topics. The trick is to

comprehend the algorithm that Twitter employs. The

positive-message buzzer team is assigned to create

trending topics with the ability to 7500 organic

accounts carried positive messages and the buzzer

team on the dark side spreading negative messages

of opposing candidates. This technique is called an

operation is known as Trending Topic Operations

(OTT: Operation Trending Topic). This research also

unfolded that the buzzer team has no standard method

or strategy for using hashtags to the target trending

topic. They only interpret the Twitter algorithm by

closely monitoring it and relying on tactile

experiences. Generally, carrying out the Trending

Topic Operation (OTT) requiring four work patterns,

namely: (1) ripple or wave system that moves based

ICAESS 2021 - The International Conference on Applied Economics and Social Science

356

on local issues, for example, a presidential candidate

is visiting a particular area, the buzzer team in that

area becomes the conversation leader on Twitter and

then followed (retweet, comment) by teams from

other regions, (2) a central system (blasting), where

this system made one narrative that moves

simultaneously, for example, the Jakarta buzzer team

sends the same narration to 7500 accounts and is

immediately blasted simultaneously by all accounts

in the same time, (3) the shift system, which is used

to maintain trending topic, where the system is carried

out in turn by each region, coordinating with the head

office in Jakarta to maintain trending, and (4)

conversation strategy, i.e., making interactions

between Twitter accounts, mentioning and greeting

each other. Through Twitter API, we successfully

gathered 328,474 Twitter accounts, with a total of

27,635 hashtags for Social Network Analysis (SNA)

a week before voting day, from 10 to April 16, 2019,

aiming to investigate the general behavior of two

presidential candidate campaigns as follows:

Figure 1: Social Analysis Networking result of two

candidacies.

From figure 1, the cluster tweet of presidential

candidate 01 (Joko Widodo - Ma'ruf Amin) supporter

is marked in red, while the presidential candidate 02

(Prabowo Subianto - Sandiaga Uno) supporter is

marked blue. Further, we analyzed the cluster

formation and connections to see the interaction

among these accounts independently.

Figure 2: Interactivity among the cluster

The above hashtag-less network visualization shows

that interaction between accounts in the presidential

candidate 02 cluster is more solid than the

presidential candidate 01. The interactions that occur

between accounts in cluster 02 tend to be more

intense. On the other hand, accounts in cluster 01 tend

to be split with the central cluster as the center and

connected to the clusters of the smaller ones. This

indicates that there are small groups in this cluster.

Small groups may form due to differences in

discussion in each small group. Our view is that

cluster 01 is formed from many driving groups

(presidential candidate social media team) on Twitter.

This research also found an interesting paradox

that significantly elevated our perspective on voters'

precincts in Indonesia. A monthly Political Indicator

Institute survey in January 2019 stated an exciting

finding that Twitter is one of the platforms rarely used

by Indonesian netizens compared to Whatsapp,

Facebook, and Instagram. If referring to the National

Survey Release of this indicator, the relation of

electability opportunities with each other's hashtag

wars of presidential candidate supporters is even

more challenging to measure. Twitter echo and

counts are still minimal (only 2% of total social media

users in Indonesia). However, the presidential

election-winning strategy doesn't always talk

numbers in quantity, but there are also quality

considerations. For the Jokowi-Amin 01 Presidential

Candidate winning team, this strategy is somewhat

successful. The selected political communication

approach is to dominate the hot issues or establishing

trending topics on Twitter. One of the reasons is that

trending topics on Twitter, which is typically a

subject of discussion on social media, generates a

Buzzer Works on Human-Machine Agency Role in Securing Indonesia’s Voters Precinct

357

resonance of information that travels across

platforms. This may happen due to the user segments

of these channels who post on other social media. In

addition to being a trending topic, interesting tweets

often become references, angles, or quotations from

traditional media such as newspapers, television,

online media, and numerous other social media.

Trending Topic Operation has taken propaganda to a

new level and been linked to the general process of

socialization. From the expert's perspective, OTT is

one of the keys to success in winning votes in the

Presidential Election, especially among millennials.

Undecided voters can also be affected by the trending

topic. The massive number of social media users,

coupled with a lack of digital literacy, further

obscures the purpose of social media; in reality, many

have a negative effect owing to the abundance of false

and untrustworthy news. Persuasion is seen as an

individual psychological process, a mass message,

and an intended audience is a large number of people

(O'Shaughnessy, 2004). Social media is an "echo

chamber" that prevents the interchange of ideas and

critical thought since users are not seeking opposing

information but rather any material that confirms their

viewpoint. This phenomenon is rather problematic

from a unity perspective.

4.3 The Agency of Change

The machine-human interactivity shown in bot and

human accounts describes the form of interaction that

widens its effect on the media. It affects the

communication psychology of its users and poses

significant consequences for knowledge, attitudes,

and user behavior in real life. The most striking

change brought by interactive features is active users

who are passive recipients of media messages. They

are empowered to perform various actions through

the media, participate in construction messages they

consume, and engage across multiple interaction

activities. This model is used as one of the strategies

of the presidential candidate support team to grab the

top spot in trending topics. Their strategy

calls it a "conversation" strategy or greets each other

between the group and systematic and measurable

arrangements that provoke other Twitter users to

respond, which is seen as natural by the Twitter

algorithm. The use of artificial intelligence of bots to

create counter-message, respond to specific issues,

and initiating twitwar, has been widely used by both

presidential candidacies. The bot will infiltrate the

opposite candidate side to interact naturally in their

cluster while assuming their essential task; to prevent

the cluster from establishing a trending topic. The

actors consisting of organic accounts (human agency)

and bots (machine agency) support each other with

various strategies that are determined in stages.

Individual organic accounts and bots create

respective threads in the form of a new narrative,

reply, retweet, and like to achieve the target goal,

namely trending topic.

5 CONCLUSIONS

In the research conducted, we concluded that the

buzzer's actions managed by the team of each

presidential candidate impact the number of voters,

especially those who have not found a choice

(undecided voters) and who still have not made a

choice (swing voters). The condition resulted in the

promising massive voter's precincts by deceiving the

Twitter algorithm. The human agency played an

essential role in creating trending topics, being unable

to stand alone, and interacting with the bot as a

machine agency to amplify the messages. Its use was

prevalent among 2019 Indonesia presidential

candidacies. Findings also showed that after the 2019

presidential election, polarization was unavoidable as

the result of the election. The stigma of cebong and

kampret is deeply rooted at the grassroots level and

hardly vanishes. Prabowo Subianto and Sandiaga

Uno, who have joined in Jokowi's cabinet, should

have been one way to cease differences. But there is

no actual movement to extinguish polarization at the

grassroots. This fact shows that the behavior of voters

who are frantically defending their decision is just a

method of play for elites, including accessing people's

brains via digital media as a buzzer. All in all, the

condition may cause polarization to persist, and it is

not inconceivable that it will become a bomb at a

critical time.

REFERENCES

Akmaliah, W. (2018). Bukan Sekedar Penggaung

(Buzzers): Media Sosial dan Transformasi Arena

Politik. MAARIF, 13(1). www.kompas.com,

Camil, R., Attamimi, N., & Esti, K. (2017). Di Balik

Fenomena Buzzer Memahami Lansekap.

Dimedjo, E. (2019). Sejarah Keberadaan Buzzer Politik di

Indonesia. Tagar.ID.

Dimitrova, D. v., Shehata, A., Strömbäck, J., & Nord, L. W.

(2014). The Effects of Digital Media on Political

Knowledge and Participation in Election Campaigns.

Communication Research, 41(1).

https://doi.org/10.1177/0093650211426004

ICAESS 2021 - The International Conference on Applied Economics and Social Science

358

Habermas, J. (1989). The Structural Transformation of the

Public Sphere: An Inquiry into a Category of Bourgeois

Society. Polity Press Cambridge.

Jhaver, S., Karpfen, Y., & Antin, J. (2018, April 21).

Algorithmic Anxiety and Coping Strategies of Airbnb

Hosts. Proceedings of the 2018 CHI Conference on

Human Factors in Computing Systems.

https://doi.org/10.1145/3173574.3173995

Jungherr, A. (2016). Twitter use in election campaigns: A

systematic literature review. Journal of Information

Technology & Politics, 13(1).

https://doi.org/10.1080/19331681.2015.1132401

Klinger, U. (2013). MASTERING THE ART OF SOCIAL

MEDIA. Information, Communication & Society,

16(5). https://doi.org/10.1080/1369118X.2013.782329

Lim, M. (2012). Clicks, Cabs, and Coffee Houses: social

media and Oppositional Movements in Egypt, 2004-

2011. Journal of Communication, 62(2).

https://doi.org/10.1111/j.1460-2466.2012.01628.x

McLuhan, M. (n.d.). This Is Marshall Mcluhan: the

Medium Is the Message (7th ed.). NBC.

Murthy, D., Bowman, S., Gross, A. J., & McGarry, M.

(2015). Do We Tweet Differently from Our Mobile

Devices? A Study of Language Differences on Mobile

and Web-Based Twitter Platforms. Journal of

Communication, 65(5).

https://doi.org/10.1111/jcom.12176

O’Shaughnessy, N. J. (2004). Politics and Propaganda:

Weapons of Mass Seduction. Manchester University

Press.

Pariser, E. (2011). The Filter Bubble: What the Internet Is

Hiding from You. Penguin UK.

Pohjonen, M., & Udapa, S. (2017). Extreme Speech Online:

An Anthropological Critique of Hate Speech Debates.

International Journal of Communication, 11.

Saraswati, M. S. (2018). Social Media and the Political

Campaign Industry in Indonesia. Jurnal Komunikasi

Ikatan Sarjana Komunikasi Indonesia, 3(1).

https://doi.org/10.25008/jkiski.v3i1.124

Setianto, Y. P. (2015). Media Policy in The Context of

Global Media Flows, The Internet, and Piracy: An

Historical Analysis of Media Regulation in Indonesia.

Jurnal Komunikasi, Malaysian Journal of

Communication, 31(2).

https://doi.org/10.17576/JKMJC-2015-3102-21

Shyam Sundar, S. (2008). The MAIN Model: A Heuristic

Approach to Understanding Technology Effects on

Credibility. 73–100.

https://doi.org/10.1162/dmal.9780262562324.073

Suler, J. (2017). The Dimensions of Cyberpsychology

Architecture. In Boundaries of Self and Reality Online:

Implications of Digitally Constructed Realities.

https://doi.org/10.1016/B978-0-12-804157-4.00001-3

Sundar, S. S., Jia, H., Waddell, T. F., & Huang, Y. (2015).

Toward a Theory of Interactive Media Effects (TIME).

In The Handbook of the Psychology of Communication

Technology. Wiley.

https://doi.org/10.1002/9781118426456.ch3

Syahputra, I. (2017). DEMOKRASI VIRTUAL DAN

PERANG SIBER DI MEDIA SOSIAL: PERSPEKTIF

NETIZEN INDONESIA. Aspikom, 3.

Tufekci, Z. (2017). Twitter and Tear Gas: The Power and

Fragility of Networked Protest. Yale University Press.

Woolley, S. C., & Howard, P. N. (2018). Computational

Propaganda (S. C. Woolley & P. N. Howard, Eds.; Vol.

1). Oxford University Press.

Buzzer Works on Human-Machine Agency Role in Securing Indonesia’s Voters Precinct

359