Who Leads? Who Follows? Exploring Agenda Setting by Media, Social Bots and Public in the Discussion of 2022 South Korea Presidential Election

Who Leads? Who Follows? Exploring Agenda Setting by Media, Social Bots and Public in the Discussion of 2022 South Korea Presidential Election | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Who Leads? Who Follows? Exploring Agenda Setting by Media, Social Bots and Public in the Discussion of 2022 South Korea Presidential Election Menghan Zhang, Ze Chen, Xinyan Liu, Jun Liu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3023846/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Social media not only changes the traditional communication environment, but also brings new changes to agenda setting. The main body of agenda setting has shifted from the traditional media to the politicians, political parties and grassroots people. With the increasing use of social bots in public opinion manipulation and political election interference, whether they can participate in or influence agenda setting has become an urgent concern. So far, there is less literature focusing on engagement in agenda-setting for social bots. This paper studies the social media discussion content of the South Korean presidential election, determines the participation of social bots, and explores the connection between media agenda, bot agenda and public agenda from the perspective of agenda setting. The study found that while the main agendas of media, social bots and the public are not the same, their agendas are relevant. In addition, the media agenda is not timely ahead of the bot agenda and the public agenda, and the time order only appears between the social bots and the public. agenda setting social media social bots Twitter political communication general election Figures Figure 1 Figure 2 1. Introduction Social media has changed the traditional communication environment and agenda setting situation[ 1 ],[ 2 ]. Social media provided an important platform to openly discuss top political and social issues[ 3 ],[ 4 ],[ 5 ]. In contrast to traditional media, the gatekeeping power on social media has reduced [ 2 ], content posted need not undergo editorial management nor scientific vetting [ 6 ], [ 7 ], which have actually expanded the number and types of actors who potentially have the ability to dominate political discourse and shape agenda. So, the influence and shaping of public agendas are no longer the exclusive power of the news media in the age of social media[ 8 ]. Political parties, politicians, grassroots and etc are able to participate in shaping the agenda, which complicates the situation of agenda setting and creates the new questions of who sets political agendas in digital platforms like social media[ 2 ]? Especially, this complexity and uncertainty becomes more prominent as powerful emerging technologies evolve. Empirical studies have confirmed that social bots supported and manipulated by algorithms[ 9 ], have appeared in many political elections and events from more than ten countries around the world [ 10 ], including the 2016 US Presidential election and the 2017 French Presidential election[ 11 ]. Social bots seem to have become a mainstream tool of public opinion manipulation, indispensable in elections and other political activities. These automated accounts on social media, deployed by the actors behind them for a certain purpose, mimic the characteristics of human user and spread various information at a very high speed, including false information and junk news[ 9 ],[ 12 ],[ 13 ],[ 14 ], which have affected the public opinion environment of social media. This has led us to wonder whether such social bots with powerful algorithm-based support can also participate in the shaping agenda? The question is the important starting point of this study. 2. Present research Agenda-setting theory has been inseparable from political elections since its inception [ 2 ]. The study by McCombs and Shaw published in 1972 showed that there was a high correlation between agendas voters considered were important and agendas reported prominently in the media, namely, the mass media influence what people think about [ 15 ]. Given the increasing number of emerging media platforms, scholars have questioned the applicability of traditional media effect models, such as the agenda setting in this complex, fragmented media environment[ 16 ]. Some studies still suggest that while the digital age provides many opportunities for ordinary individuals and other organizations to spread information, their impact on public opinion is limited. On social media, traditional media, such as newspapers, would still publish online news by registering their official accounts. While other information on social media somewhat weakens media attention, media still plays a critical role in agenda setting [ 8 ]. By analyzing tweets about Paris attack, researchers found that professional mass media organizations still hold a greater agenda-setting power than individual opinion leaders for setting the public agenda, as they obtain significantly more tweets, mentions, and replies from the public [ 17 ]. However, some researchers think that agenda setting needs to be relooked into as the public agenda is taking a central stage through the new/social media in the digital and globalised era [ 18 ]. Meraz made a similar point after studying the blog and social media agenda setting. She believed the popularity of independent blog networks had allowed citizens more influence and power in setting news agendas[ 19 ]. In addition, some actors with a significant resources, such as political parties, politicians have the opportunity to participate in the discussion and agenda setting on social media. A president like Donald Trump, who can used social media to lie to the public, had also agenda-setting effects [ 20 ]. Furthermore, social bots, algorithm-supported, automated tools deployed on social media, are also fighting for salience on political issues and attributes[ 46 ] (Vargo, 2018). The distinguishing feature of social bots is that they can systematically and repeatedly release large amounts of information in a short period of time and connect with more users in the social network[ 9 ]. In general elections, social bots are often used to intervene in political discussions. For example, through orchestrating, to create the impression of an organic support for political actors with the help of social bots, or to systematically discredit the candidate and weaken his reputation [ 11 ]. As the emerging political opinion climate maker, social bots not only disturbed the public opinion environment, but also brought new challenges to the traditional agenda setting situation[ 12 ]. On March 10th, 2022, the 20th South Korea presidential election ended with Yoon Seok-yeol’ victory margin of 0.73%, which is the narrowest margin of victory in any presidential election in South Korea. In fact, the 20th South Korea Presidential election has been unusually intense since the primary election. Several opinion polling since 2022 have shown that two main candidates, Lee Jae-myung and Yoon Seok-yeol, have received very close approval ratings, presenting an evenly matched situation. Under the situation, any party with significant resources would all probably deploy state of the art technologies to enact influence operations and other forms of manipulation of public opinion [ 11 ]. If social bots were also deployed in debate of the South Korea presidential election, what impact would they have on the political discussion? But for present, this suppose is not determined. A lot of research focuses on social media and political elections, with more research on predicting election results through social media data[ 21 ]. Or studying the impact of social media on the election, including the negative effects and prevention measures[ 22 ]. Some researchers also focus on the changes in the agenda setting on digital platforms, including the characteristics and changes of the media and other actors participating in the political agenda setting [ 23 ] ,[ 24 ], [ 25 ]. The research focusing on the participation of the social bots in the political election mostly focuses on the characteristics of the bots and the impact on the public opinion environment generated by bots[ 26 ],[ 27 ],[ 28 ]. Yet, we believe that social bots, which are greatly enhancing as the increasing sophistication of Artificial Intelligence[ 11 ], also have the potential to participate in agenda setting. And what would be the contact between the bot agenda, media agenda and public agenda in the complex ecosystem like social media? In traditional agenda-setting, the mass media determines what the public thinks about and how to think[ 15 ], that is, the media agenda takes precedence over the public agenda. So would the bot agenda also be influenced by the media agenda? And would the agenda by bots influence the public's perception of the relevant agenda? Therefore, this work takes the South Korea presidential elections as a case, based on determining the involvement of social bots in political discussions, to explore the contact between the media agenda, bot agenda and public agenda from the perspective of agenda setting. The specific research questions are as follows: RQ1: What are the priorities of agenda perceived by media, bot and public? RQ2: Is there correlations between the media agenda, bot agenda, and public agenda? RQ3: Media agenda, bot agenda, and public agenda who takes the lead in time? 3. Methodology This study used the following data and methods to achieve our research objectives. First, tweets related to 2022 South Korea presidential elections were the material used for analysis. Second, we used Botometer to identify social bots, and divided all accounts into media accounts, social bots and human users. Third, to address the RQ1 of this study, cluster analysis would be used to cluster the tweets and the agendas included in the discussion of 2022 South Korea presidential elections would be determined. And the manual and computer-aided topic coding were used to determine the distribution of the agenda in media, bot and public accounts. Fourth, to solve the RQ2, we would analyze the correlation of media agenda, bot agenda, and public agenda using SPSS software. Fifth, to solve the RQ3, we used Almon Polynomial Lag to identify the time lag effect. Through the result of time lag, whose agenda was ahead in time would be judged. We introduce the way of data collection and the specific application of these methods in the following paragraphs. 3.1 Data Collection Discussions related to 2022 South Korea presidential elections on Twitter were taken as the contents of the analysis in the study. Choosing Twitter as the data collection platform mainly considers the following reasons: (1) Twitter is currently the most popular tool adopted by news organization for content dissemination. (2) Twitter is one of the most-used social media by South Korea Internet users[ 29 ]. (3) Currently, social bots are the most deployed on Twitter[ 9 ]. Data collection was conducted from February 15 to March 9, 2022. The two timestamps are very important time points in the South Korea Presidential elections. On February 15, the South Korea Presidential elections officially entered the election campaign stage, which means that the candidates will conduct a mass publicity and canvassing campaign. Voting for the Presidential elections officially ended on March 9. Meanwhile, given that this work targeted Twitter users from South Korea and their tweets, we only collected tweets in the language of Korean. Hashtags were used to search for related tweets. On Twitter, hashtags are often added “ # ” thus aid the formation specific themes and topics[ 30 ]. Before data collection, we identified a list of hashtags through the following steps. First, we screened out the hashtags related to the South Korean presidential election from the top twitter trending hashtags between February 15 and March 9,2022. After that, combining the words in the headlines of major South Korea newspapers, we perfected the hashtags list. Last, we sent this hashtags list to five native Korean Twitter users aged between 20–40 years for confirmation and manual supplementation[ 31 ]. The six hashtags appearing in Table 1 were finally confirmed and used. We collected all the tweets covered in these tags. The official API data interface service provided by Twitter was used to collect tweets. When collecting data, the privacy of Twitter users is respected without any personal information collected and displayed. Table 1 Hashtags were used to search for related tweets. Hashtags Meanings #제20대대통령선거 The 2022 presidential election #대선 General election #윤석열 Yoon Seok-yeol #국민의힘 People Power Party #이재명 Lee Jae-myung #더불어민주당 Democratic Party of Korea 3.2 Detection of Social Bots Currently, there are many methods to identify social bots, including crowd sourcing-based recognition methods, identification methods based on social network information, and machine learning-based recognition methods [ 32 ]. This work uses a machine learning-based recognition method, whose representative tool is Botometer, developed by Institute of Network Science of Indiana University and open to all users, which is a typical framework for machine recognition of social bots[ 33 ]. The API calling method provided by Botometer pro was used to detect a large number of accounts. Botometer returns a bot score by extracting and analyzing more than 1,000 features, including users’ tweeting frequency, behavioral characteristics, personal profiles, social networks, and friends’ characteristics[ 34 ]. Scores closer to 1 represent a higher chance of being bots, while those closer to 0 are more likely to belong to humans[ 13 ]. Yet, at present, there is not agreement on a threshold that can reliably distinguish bots from humans. On the threshold setting, we want to refer to the complete automation probability (CPA) officially provided by Botometer and the studies had published. If an account has a higher score than the defined threshold, we will defined as a bot. It's also worth noting that Botometer fails to output a score when encountering account suspensions and authorization problems [ 35 ], [ 13 ]. 3.3 Text Cluster Analysis based on KH Coder Software In the domain of natural language processing, clustering is an unsupervised process which is based on similarity between data to form some pattern [ 36 ]. In this process, objects that are similar to each other are assigned a specific group or class from those that are dissimilar to them [ 37 ]. Various automation software and methods including K-Means, Affinity Propagation and others have been posed for clustering. For clustering analysis of texts, this work used KH Coder, an open source software for quantitative content analysis, text mining and computational linguistics[ 38 ]. Based on the clustering large applications (CLARA), a relatively large number of documents are able to process by KH Coder. Moreover, with the support of the CLARA algorithm, the KH Coder can realize the comparison of the sample clustering, and select the optimal center clustering [ 38 ]. Importantly, multiple natural languages, including Korean, can be recognized and processed by KH Coder. Prior to performing the text clustering analysis, we first input a copy of Korean stop words into KH Coder, when processing it can automatically filters stop words out. After processing the tweets, KH Coder can return optimal clusters, each containing high probability words[ 38 ]. Based on the clustering results and the content of the original tweets, we would name each cluster returned by the KH coder, forming agendas of significant meaning. Later, each tweet was coded based on agendas by manual coding and computer-aided analysis, and the distribution of the agenda in media, bot and public accounts would be determined. 3.4 Correlation Analysis based on SPSS Correlation analysis was tested for whether there was a significant relationship between the agendas of the two actors. We would analyze the correlation of media agenda, bot agenda, and public agenda using SPSS software. Media agenda, bot agenda, and public agenda were grouped pairwise and analyzed separately for correlation. The correlation of two types of agenda, such as the correlation between the media agenda and bot agenda, is judged by the hourly number of tweets posted under agendas. Firstly, the Kolmogorov-Smirnov test (KS text) was used to assess normality for data [ 39 ]. KS test results showed that the data did not fit the normal distribution. So, Spearman correlation was used to determine the relationship. The significance level was set at p˂ 0.05 and the strong relationships set at p˂ 0.01 [ 39 ]. 3.5 Time-Lag Analysis based on the Almon Polynomial Lag Almon Polynomial Lag is a novel technique for estimating the weights of a distributed lag by means of a polynomial specification by Shirley·Almon[ 40 ]. It transforms multiple jet-lag variables into a polynomial form that minimizes the multicollinearity problem and has the advantage of reducing the loss of free degree [ 40 ] Since its introduction, the Almon lag technique has been widely used in empirical work. The main reason for its popularity is probably the ease with which it can be used-simply pick a length of lag, and a degree of polynomial, and results are quickly forthcoming [ 41 ]. At present, the method is applied not only in econometrics and political economy, but also in social media data analysis[ 42 ]. Eviews software was used for analysis in this study and the analysis process as follows. Firstly, determining the objects of the time-lag analysis. The objects of the time-lag analysis were determined based on the results of the correlation analysis. Two types of agendas with significant correlations would be the object of the time-lag analysis. Combinations without correlations were not objects of the time-lag analysis. Secondly, determining the time units. The previous studies, used time series methods for agenda lag analysis, were set in years, month or days [ 43 ],[ 44 ]. However, in this study, we performed the analysis of time lags in hours. Two reasons are specifically considered. One is the time range of the data selected in our study is one month, the other is the information spreads fast on social media, especially, the number of topics and tweets would increase near the presidential election. The discrepancy in time between different agendas is more obvious by hours. Finally, presenting the analysis results. After entering data normatively, the software could return the result of the lag order. The orders shown by results are the hours between the two agendas. For example, when x = bot, y = human, the result is fourth-order, which indicates that the bot agenda lags behind the public agenda by order 4, namely four hours. 4. Results As mentioned above, we obtained 45,169 tweets under six hashtags from February 15 to March 9,2022. The language of all the tweets is in Korean. To facilitate the subsequent data processing, repeated tweets, hyperlinks, and emoji symbols were filtered. After data cleaning, a total of 44,009 valid tweets posted by 18,700 users. According to the research approach of this work, after detecting social bots and dividing the three types of accounts, we conducted cluster analysis and time lag analysis successively, and obtained the corresponding results. 4.1 Social Bots Detection Results Through a specific Python program, 18,700 Twitter users were entered to Botometer to obtain scores. The final scores of 17,362 accounts were returned, and 1,338 accounts could not be identified and detected by Botometer. So, the accounts and tweets that cannot be identified and detected by Botometer not be considered as the subjects of this study. 17,362 Twitter accounts and the 42,076 tweets they posted constitute the dataset of this study. We set the threshold to 0.8 to separate humans from bots. In other words, accounts with a score greater than 0.8 are bots, and accounts with less than 0.8 are human and media. Followed, we manually classified and counted the accounts of bots, humans and medias. There were 7,490 social bots, accounting for 43.14% of the total users in the dataset, who produced 18,206 tweets. 19,835 tweets were posted by 9,834 human accounts. And only 38 media accounts certified by Twitter, in the dataset, who published 4,035 tweets. Figure 1 presents the number of the three types of accounts and their tweets. 4.2 Text Clustering Analysis and Agenda Distribution Results After processing the tweets, KH Coder returned 13 clusters. And 12 lists of words of practical meaning were marked with names, forming 12 agendas, which are displayed in Table 2 . Because cluster 13 has no interpretable meaning, it was not presented. The 12 agendas can be divided into two categories. The one category is directly related to the South Korea general election, including economic resurgence, political liquidation, social welfare, Democratic Party of Korea, People Power Party and stigmatization propaganda. The other category is not directly related to the general election, including international relation, Russia-Ukraine conflict, gender issues, Seoul real estate, fasting dog meat and forest fire. Table 2 List of 12 agendas. Agendas High Probability Words International relation 대한민국(South Korea), 미국(U.S.A), 중국(China), 일본(Japan), 북한에(Democratic People’s Republic of Korea), 캐나다(Canada), 호주(Australia), 터키(Turkey), 인도(India), 파키스탄(Pakistan), 러시아(Russia), 우크라이나(Ukraine) Russia-Ukraine conflict 대한민국(South Korea), 러시아(Russia), 우크라이나(Ukraine),전쟁(war), 평화(peace), 관계를(relationship), 대통령이(president), 푸틴(Vladimir Putin), 젤렌스키(Zelensky), 나토는(the North Atlantic Treaty Organization), 충돌(conflict), 핵(nuclear weapon), 공격을( attack), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeol) Economic resurgence 민생경제(livehood economy), 경제(economy), 회복을(recovery), 노동자(worker), 최저임금(minimum wage), 최저임금제도를(Minimum wage system), 알바(part-time job), 자영업자(individual household ), 중소상공인(middle and small industrialist and merchant), 부담을(burden), 부채(be in debt), 대출(loan),저소득(low-income) Political liquidation 대한민국(South Korea), 대선(general election), 대통령을 (president), 후보 (candidate),정권교체(regime change), 신천지(new field ), 정권(regime), 교체(replace), 전환(transition), 리셋 (reset),문재인(Moon Jae-in), 이재명(Lee Jae-myung), 더불어민주당(Democratic Party of Korea),국민의힘(People Power Party), 윤석열(Yoon Seok-yeol) Social welfare 복지국가(welfare state ), 복지(welfare), 노인(elder), 노인요양(The old man recuperation), 장애인돌봄(Care for the disabled), 건강보험(healtb-insurance), 휴게수당(recess allowance), 초등돌봄(Primary care), 아동(children), 수당(allowance), 보육(child welfare), 육아(child rearing),국가장학금(National scholarship), 대학생(college student) Gender issues 남성(man), 여자(woman), 성질(sexual), 양성평등(sex equality),여성(gender), 성별(female), 주부(housewife),갈등(contradiction), 화합을(harmony), 혐오(disgust), 범죄(crime), 성범죄(sexual crime), 문제(problem),여성가족부를(Women's family department),페미니즘은(feminism),후보 (candidate) Seoul Real Estate 서울(Seoul), 부동산(real estate), 집값(housing price), 폭락(steep fall), 폭등(spurt in prices), 청년(youth), 내집마련(buy a house), 청년층(Youth order), 좌절감이(frustration), 불공정(unfair), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeo) Fasting dog meat 대한민국(South Korea), 개고기(Dog meat), 유통시장(trading market), 모란시장(Peony market), 동물(animal), 반려동물 (pet), 법률(law), 금지(ban), 전통(tradition), 지원(support), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeol) Forest Fire 대한민국(South Korea), 강원도(Gangwon-do),산불(wildfire), 삼림(forest), 산림( mountain forest), 화재(fire), 소방청(Fire hall), 소방대원(Firefighters), 이재민(victims of a natural calamity), 원자력(Nuclear), 발전소(Power Plant),현장(scene), 피해(lose), 대선(general election), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeol) Democratic Party of Korea 대한민국(South Korea), 대선(general election),더불어민주당(Democratic Party of Korea),후보가(candidate), 이재명(Lee Jae-myung),기호1번(No.1), 1번(No.1), 1번남(No.1 male),투표(vote), 문재인(Moon Jae-in), 지지율(support rate), 실현(support,) 나를위해(for me),외롭지(lonely) People Power Party 대한민국(South Korea), 국민의힘(People Power Party), 후보가(candidate), 윤석열(Yoon Seok-yeol), 지지율(support rate), 2번남(No.2 male), 기호2번(No.2),투표(vote), 낙선이고(lose an election), 터무니없다(absurd) Stigmatization Propaganda 흑색선전(stigmatization propaganda),이재명(Lee Jae-myung), 개발사업(developmental project), 경기도(Gyeonggi), 대장동(),부동산(real estate), 아들(son), 성매매(Sex deals), 도박(gamble), 성매수(go whoring), 윤석열(Yoon Seok-yeol), 성접대(Sexual entertainment), 김건희(Kim Keon-hee), 주가조작(Manipulation of stock price), 통정매매(Overall trading), 거짓말(lie), 경력(qualifications), 학력(educational background), 위조에(counterfeit), 장모님(wife's mother) Then, we analyzed the overall data with manual coding and Python program-assisted query, i.e., topic-coding for each tweet. When a tweet matches one or more keywords on an agenda, it is marked 1 by the program, and 0 when there is no match. Followed, we did the statistics for the number of tweets posted by three types of accounts under the 12 agendas after obtaining the encoding results to determine the distribution proportion of the agenda in media, bot and public accounts. Figure 2 presents the results. It can be seen that the media agenda did not involve in international relations. The bot agenda was more focused on economic resurgence, political liquidation, Democratic Party of Korea, People Power Party and stigmatization propaganda, and the public agenda was more focused on international relations, Russia-Ukraine conflict, gender issues, social welfare, Seoul real estate, fasting dog meat, and forest fire. 4.3 Correlation Analysis Results Table 3 presents the bivariate correlation coefficients under the 12 agendas and their significance results.The results show that the media agenda and bot agenda have significant correlation in social welfare, gender issues, Seoul real estate, fasting dog meat, Democratic Party of Korea, People Power Party, and stigmatization propaganda. Bot agenda and public agenda have significant correlation in the Russia-Ukraine conflict, economic recovery, political liquidation, social welfare, fasting dog meat, forest fire, Democratic Party of Korea and stigmatization propaganda. Media agenda and public agenda have a significant correlation in gender issues, economic resurgence, social welfare, Seoul real estate, fasting dog meat, and stigmatization propaganda. Table 3 Results of the correlation analysis. Agendas Bivariate Correlation Coefficent p value Results International relation Media-Bot / / / Bot-Human .003 .978 Uncorrelation Media-Human / / / Russia-Ukraine conflict Media-Bot − .116 .068 Uncorrelation Bot-Human .172** .006 Uncorrelation Media-Human − .108 .087 Uncorrelation Economic resurgence Media-Bot − .082 .361 Uncorrelation Bot-Human .242** .007 Correlation Media-Human − .192* .032 Correlation Political liquidation Media-Bot − .011 .819 Uncorrelation Bot-Human .124** .009 Correlation Media-Human .044 .352 Uncorrelation Social welfare Media-Bot − .147* .018 Correlation Bot-Human − .315** .000 Correlation Media-Human − .299** .000 Correlation Gender issues Media-Bot − .309** .000 Correlation Bot-Human .039 .647 Uncorrelation Media-Human − .246** .003 Correlation Seoul Real Estate Media-Bot − .386** .000 Correlation Bot-Human − .024 .682 Uncorrelation Media-Human − .354** .000 Correlation Fasting dog meat Media-Bot − .338** .000 Correlation Bot-Human .367** .000 Correlation Media-Human − .346** .000 Correlation Forest Fire Media-Bot .009 .894 Uncorrelation Bot-Human .532** .000 Correlation Media-Human .031 .65 Uncorrelation Democratic Party of Korea Media-Bot .271** .000 Correlation Bot-Human .176** .000 Correlation Media-Human .075 .101 Uncorrelation People Power Party Media-Bot .121** .009 Correlation Bot-Human .06 .192 Uncorrelation Media-Human .049 .29 Uncorrelation Stigmatization propaganda Media-Bot .189** .000 Correlation Bot-Human .185** .000 Correlation Media-Human .105* .026 Correlation 4.5 Results of the Time-Lag Analysis Table 4 is the results of the time-lag analysis based on Almon polynomial. Overall, the time lags exist mainly between bot agenda and public agenda. Bot agenda lag behind public agenda in Russia-Ukraine conflict, fasting dog meat and forest fires. Public agenda lag behind bot agenda in economic resurgence, political liquidation, Democratic Party of Korea and stigmatization propaganda. Yet, the time lags does not exist between the media agenda and the other two types of agendas. Table 4 Results of the time-lag analysis. Agendas Time-lag (x = bot, y = human) Time-lag (x = human, y = bot) Russia-Ukraine conflict fourth-order / Economic resurgence / sixth-order Political liquidation / first-order Fasting dog meat ninth-order / Forest fires sixth-order / Democratic Party of Korea / third-order Stigmatization propaganda / third-order 5. Discussion Through exploring, we found that there were indeed social bots involved in election discussions on Twitter during the South Korea election. We would discuss and conclusion based on the results of this study. Firstly, the priorities of the media agenda, bot agenda, and public agenda are different. On the agendas of social welfare, stigmatizing propaganda and the People Power Party, the proportion of media tweets is higher than on other agendas. Moreover, the media did not post tweets on social media related to the agenda of international relations. That is to say medias usually focused the agenda setting on domestic issues rather than diplomatic issues when faced the major domestic political events. Although the bot agenda was also directly related to the presidential election, the content was more diverse compared with media. We also found that the bot agenda exhibited characteristics by regularity. They set the agenda by heavily forwarding existing news stories and messages on social media. So, they were active on some agendas, such as the agenda of stigmatization propaganda involved the candidates' past disgraceful events, and the agenda of political liquidation that has occurred in successive presidential elections. The public agenda involves more content than the media agenda and bot agenda. Unlike social bots, the public tends to express their viewpoint and opinions on general election-related agendas such as economic resurgence, political liquidation, and social welfare. And, they are very sensitive to hot issues and pay close attention to current events such as the Russia-Ukraine conflict and the forest fires in South Korea. Secondly, we found that the agendas of the three type of actors are not isolated, the correlation results show that there are still significant relationships on part of the agenda between media and social bots, between media and the public, and between social bots and the public. For example, in the agenda of the fasting dog meat, there is a correlation between media and bots, between media and public, and between bots and the public. Although the issue is not directly related to the election itself, it is highly controversial in South Korea. Different attitudes toward "eating dog meat" had also become a way for candidates to gain support. It shows that controversial issue are something that media, bots and humans all pay attention to. And in the process of participating in the discussion, they have the situation of quoting and forwarding each other. Thirdly, the temporal order of the agendas of the three types of actors were presented through a time-lag analysis. The results showed that the media agenda was not ahead of the bot agenda and the public agenda in time, and that the time order only appeared between social bots and the public. That said, the media agenda does not lead the bot agenda and public agenda. We believe that the media agenda was not ahead of the public agenda and was related to the public distrust of the information disseminated by the media, especially on social media[ 45 ]. The reason why the media agenda is not ahead of the bot agenda may be that social bots tend to interact with the general public rather than with the media in this election discussion. Social bots lead the public on four agendas: economic resurgence, political liquidation, democratic Party of Korea, and stigmatization propaganda.In the face of these issues, social bots simply spread information through massive retweets, while humans published their own ideas and produced information based on specific issues. After spending a certain "reaction time" and "buffer time", people naturally lag behind bots, the automated tools for disseminating information based on algorithms. The public agendas lead bot agendas in Russia-Ukraine conflict, fasting dog meat and forest fires. The three agendas are in not directly related to the South Korea presidential election itself. But the people's discussions have still linked them to the presidential election. The forest fire in Gangwon-do during the election was seen by some superstitious people as a symbol of unlucky. As candidates Lee Jae-myung and Yoon Seok-yeol have different attitudes to fasting dog meat, people discussed the topic while expressing their support or not attitude for two candidates. Social bots operated mechanically that had not advantage on the discussion of these active agenda. 6. Conclusion In the digital age, the development of various emerging platforms and the use of new technologies complicate the communication environment. Many researchers believe that although the traditional news media was central in the agenda setting in the past, the agenda setting of the traditional media is is now just one force among many competing influences in the Internet environment [ 19 ]. Multiple actors, including politicians, the general public and technical tools represented by social bots, all fight for salience on political agendas and attributes [ 46 ]. This work precisely focuses on and discusses the agenda-setting characteristics of media, social bots, and the public in the complex ecosystem of social media. Using a scientific approach, we first found the agenda used and avoided by three types of actors when facing the same event. While their agenda focus are different, they are relevant in some ways. On controversial agendas such as fasting dog meat, for example, there is a pairwise correlation between media, social bots and the public. On the basis of correlation, we further explore who takes the lead in time. We found that the media agenda was not ahead of the bot agenda and the public agenda in time. This suggests that the agenda setting monopoly power of the news media is somewhat weakened. Overall, our work reveals the dynamic features of multi-participant agenda settings on social media ecosystem where humans and bots to coexist, providing a viable path to parsing the complexity and diversity of agenda settings in digital environments. Of course, there are also some limitations of this study. First, this work mainly focuses on the South Korea election as a case study, and whether the findings we reported generalize to other countries remains to be considered, which is also one of our future research directions. Second, although our results were obtained by a scientific analysis of the data collected, our data volume is relatively small. At the same time, we will explore better ways and technologies to obtain data in the future without violating the rules of the platform. Declarations Acknowledgements Not applicable. Author contributions Menghan Zhang is responsible for researching architectural ideas, proposing research hypotheses, and analyzing the content. Ze Chen is responsible for collecting and analyzing preliminary data. Xinyan Liu is responsible for language proofreading of the research content. Jun Liu is responsible for implementing and liaising with the research, including coordination and content creation. Funding This study did not receive any funding in any form. Availability of data and materials The datasets generated and/or analysed during the current study are not publicly available due to ethical issues but are available from the corresponding author on reasonable request. Ethics approval and consent to participate This manuscript is not under review elsewhere and the results have not been published previously or accepted for publication. This manuscript has been seen and approved by all authors. All methods were performed in accordance with the relevant guidelines and regulations. The questionnaire and methodology for this study was approved by the research ethics committee of the Soochow University and University of Copenhagen before data collection. Consent for publication Not applicable. Informed consent Informed consent was obtained from all participants included in the study. Competing interests The authors declare that they have no competing interests. References Yang, X., Chen, B.-C., Maity, M., Ferrara, E.: Social politics: agenda setting and political communication on social media. In: International Conference on Social Informatics, pp. 330–344. Springer (2016) Gilardi, F., Gessler, T., Kubli, M., & Müller, S. (2022). Social media and political agenda setting. Political Communication, 39(1), 39-60. Carlisle, J.E., Patton, R.C.: Is social media changing how we understand political engagement? an analysis of facebook and the 2008 presidential election. Polit. Res. Q. 66(4), 883–895 (2013) Varol, O., Ferrara, E., Ogan, C.L., Menczer, F., Flammini, A.: Evolution of online user behavior during a social upheaval. In: Proceedings 2014 ACM Conference on Web Science, pp. 81–90 (2014) Yang, X., Chen, B.-C., Maity, M., Ferrara, E.: Social politics: agenda setting and political communication on social media. In: International Conference on Social Informatics, pp. 330–344. Springer (2016) Charles‐Smith, L.E.; Reynolds, T.L.; Cameron, M.A.; Conway, M.; Lau, E.H.Y.; Olsen, J.M.; Pavlin, J.A.; Shigematsu, M.; Strei‐chert, L.; Suda, K.J.; et al. Using Social Media for Actionable Disease Surveillance and Outbreak Management: A Systematic Literature Review. PLoS ONE 2015, 10, e0139701. https://doi.org/10.1371/journal.pone.0139701. Meleo‐Erwin, Z.; Basch, C.; MacLean, S.A.; Scheibner, C.; Cadorett, V. “To each his own”: Discussions of vaccine decision‐making in top parenting blogs. Hum. Vaccines Immunother. 2017, 13, 1895–1901. https://doi.org/10.1080/21645515.2017.1321182. Hindman, D.B., 2009. Mass media flow and differential distribution of politically disputed beliefs: the belief gap hypothesis. Journal. Mass Commun. Q. 86 (4), 790–808. Ferrara, E.; Varol, O.; Davis, C.; Menczer, F.; Flammini, A. The rise of social bots. Commun. ACM 2016, 59, 96–104.https://doi.org/10.1145/2818717. Woolley S C, Howard P N. Political communication, computational propaganda, and autonomous agents: Introduction[J]. International Journal of Communication, 2016, 10. Ferrara, E. (2020). Bots, elections, and social media: a brief overview. Disinformation, Misinformation, and Fake News in Social Media, 95-114. Kim, A. Nearly Half of the Twitter Accounts Discussing ‘Reopening America’ May Be Bots, Researchers Say. CNN. 22 May 2020. Available online: Shao, C.; Ciampaglia, G.L.; Varol, O.; Yang, K.‐C.; Flammini, A.; Menczer, F. The spread of low‐credibility content by socialbots. Nat. Commun. 2018, 9, 4787. Zhang, M., Chen, Z., Qi, X., & Liu, J. (2022). Could Social Bots’ Sentiment Engagement Shape Humans’ Sentiment on COVID-19 Vaccine Discussion on Twitter?. Sustainability, 14(9), 5566. McCombs, M. E., & Shaw, D. L. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36(2), 176–187. https://doi.org/10.1086/267990. Guo, L., & McCombs, M. (2011, May). Network agenda setting: A third level of media effects. In annual conference of the International Communication Association, Boston, MA. Yang, & Sun. (2021). Who has set whose agenda on social media? a dynamic social network analysis of tweets on paris attack. Communication Quarterly. Salman, A., Mustaffa, N., Mohd Salleh, M. A., & Ali, M. N. S. (2016). Social media and agenda setting: Implications on political agenda. Jurnal Komunikasi Malaysian Journal Of Communication, 32(1), 401-414. Meraz, S. (2009). Is there an elite hold? Traditional media to social media agenda setting influence in blog networks. Journal of computer-mediated communication, 14(3), 682-707. Skewes, E. (2018). Time delays are not enough; Media must call out lies. Journal of Media Ethics, 33(2), 97–99. https://doi.org/10.1080/23736992.2018.1435498 Cameron, M. P., Barrett, P., & Stewardson, B. (2016). Can social media predict election results? Evidence from New Zealand. Journal of Political Marketing, 15(4), 416-432. Aral, S., & Eckles, D. (2019). Protecting elections from social media manipulation. Science, 365(6456), 858-861. Metaxas, P. T., & Mustafaraj, E. (2012). Social media and the elections. Science, 338(6106), 472-473. Salman, A., Nguyen, C., Lee, Y. H., & Cooksey-James, T. (2016). A review of barriers to minorities’ participation in cancer clinical trials: implications for future cancer research. Journal of immigrant and minority health, 18(2), 447-453. Langer, A. I., & Gruber, J. B. (2021). Political agenda setting in the hybrid media system: Why legacy media still matter a great deal. The International Journal of Press/Politics, 26(2), 313-340. Hagen L, Neely S , Keller T E , et al. Rise of the Machines? Examining the Influence of Social Bots on a Political Discussion Network[J]. Social Science Computer Review, 2020(10):089443932090819. Schäfer, F., Evert, S., & Heinrich, P. (2017). Japan's 2014 general election: Political bots, right-wing internet activism, and prime minister Shinzō Abe's hidden nationalist agenda. Big data, 5(4), 294-309. Pastor-Galindo, J., Zago, M., Nespoli, P., Bernal, S. L., Celdrán, A. H., Pérez, M. G., ... & Mármol, F. G. (2020). Spotting political social bots in Twitter: A use case of the 2019 Spanish general election. IEEE Transactions on Network and Service Management, 17(4), 2156-2170. Published by Nina Jobst, Apr 11, 2022 Social media usage in South Korea - Statistics & Facts.https://www.statista.com/topics/5274/social-media-usage-in-south-korea/ Bruns, A., & Burgess, J. (2011). The use of Twitter hashtags in the formation of ad hoc publics. In Proceedings of the 6th European consortium for political research (ECPR) general conference 2011 (pp. 1-9). The European Consortium for Political Research (ECPR). Badawy, A.; Lerman, K.; Ferrara, E. Who Falls for Online Political Manipulation? In Proceedings of the Companion Proceedingsof the 2019 World Wide Web Conference, Association for Computing Machinery, San Francisco, CA, USA, 13–17 May 2019; pp.162–168. Ping, H. , & Qin, S. . (2018). A Social Bots Detection Model Based on Deep Learning Algorithm. 2018 IEEE 18th International Conference on Communication Technology (ICCT). IEEE. Luceri, L.; Deb, A.; Giordano, S.; Ferrara, E. Evolution of bot and human behavior during elections. First Monday 2019, 24, 1–12.https://doi.org/10.5210/fm.v24i9.10213. Davis, C.A.; Varol, O.; Ferrara, E.; Flammini, A.; Menczer, F. BotOrNot: A system to evaluate social bots. In Proceedings of theProceedings of the 25th International Conference Companion on World Wide Web, Montreal, QC, Canada, 11–15 April 2016;pp. 273–274. https://doi.org/10.1145/2872518.2889302. Ferrara, E. Measuring Social Spam and the Effect of Bots on Information Diffusion in Social Media. In Complex Spreading Phe‐nomena in Social Systems; Springer: Berlin/Heidelberg, Germany, 2018; pp. 229–255. https://doi.org/10.1007/978‐3‐319‐77332‐2_13. Ahuja, S. , & Dubey, G. . (2017). Clustering and sentiment analysis on Twitter data. 2017 2nd International Conference on Telecommunication and Networks (TEL-NET). IEEE. J. C. Dunn, “A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters,” Journal of Cybernetics 1973, 3: 32-57.1080/13501763.2018.1458890 Higuchi, K. (2016). KH Coder 3 reference manual. Kioto (Japan): Ritsumeikan University. Hinton, P., McMurray, I., & Brownlow, C. (2014). SPSS explained. Routledge. Almon, S. (1965). The distributed lag between capital appropriations and expenditures. Econometrica: Journal of the Econometric Society, 178-196. Peter Schmidt & Roger N. Waud (1973) The Almon Lag Technique and the Monetary versus Fiscal Policy Debate, Journal of the American Statistical Association, 68:341, 11-19 Lehrer, S., Xie, T., & Zeng, T. (2021). Does high-frequency social media data improve forecasts of low-frequency consumer confidence measures?. Journal of Financial Econometrics, 19(5), 910-933. Dearing, James W., and Everett M. Rogers (1988) "The agenda-setting process for the issue of AIDS." Paper presented to the Mass Communication Division, International Communication Association, New Orleans, 29 May-3 June. Smith, Kim A. (1987) "Newspaper coverage and public concern about community issues: A time-series analysis." Journalism Monographs 101:1-32. Smith, Ted J., Ill, and J. Michael Hogan (1987) Se-Uk Oh So-Eun Lee, 2021 Digital News Report https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2021/south-korea Vargo, C. J. (2018). Fifty years of agenda-setting research: New directions and challenges for the theory. The Agenda Setting Journal, 2(2), 105-123. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3023846","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":207890531,"identity":"6e79df50-6434-4d90-8e5a-e6bf6f0ccc4b","order_by":0,"name":"Menghan Zhang","email":"","orcid":"","institution":"Soochow University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Menghan","middleName":"","lastName":"Zhang","suffix":""},{"id":207890532,"identity":"43109716-eb3a-416c-9fd6-30343e706883","order_by":1,"name":"Ze Chen","email":"","orcid":"","institution":"Soochow University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ze","middleName":"","lastName":"Chen","suffix":""},{"id":207890533,"identity":"b620b59f-511d-4b46-97cd-25d915e09aa2","order_by":2,"name":"Xinyan Liu","email":"","orcid":"","institution":"Soochow University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xinyan","middleName":"","lastName":"Liu","suffix":""},{"id":207890534,"identity":"6851cbfe-a2f6-42c2-9692-96b7f5ca6808","order_by":3,"name":"Jun Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5klEQVRIie3QMQrCMBSA4ZQHcal0jUM9QyCgi+hVEgqOLi7dLAh107XeoiA4P8ng0gMUdBCEzgUPoGlHkVQ3h/yQIZCPJI8Ql+sPCwAQJX+uArNBs7ykiww2qcI6BjlIgCB+Q3hRiFNWgOQI7TXdhJSS635KF+KyO+maTMIcobrZhJdJaYi/HF1187C5yJGOuY0Ak2gI845l1BCtcvQpsxHKVGII9w5ZS57dxPc1Md+XKmctwW7CeikxQ0bByohjwSOx13RkJTMdPGoz4WGQqXsdx9Nwe15XVvJWMyr44bzL5XK5PvcCWuxVrOJUW9kAAAAASUVORK5CYII=","orcid":"","institution":"University of Copenhagen","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Jun","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2023-06-05 10:00:57","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3023846/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3023846/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":38299326,"identity":"0f531187-3398-4019-aa89-05613d144544","added_by":"auto","created_at":"2023-06-09 15:57:15","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":51854,"visible":true,"origin":"","legend":"\u003cp\u003eNumber of the three types of accounts and their tweets.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-3023846/v1/444a78440b2593f384541e64.png"},{"id":38299325,"identity":"f03ecd37-3b94-4312-bae7-97c9b7ffcd49","added_by":"auto","created_at":"2023-06-09 15:57:15","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":89129,"visible":true,"origin":"","legend":"\u003cp\u003eThe proportion of the agendas in media, bot and public accounts\u003c/p\u003e","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3023846/v1/8b9751ad8c9659bd7277552c.jpg"},{"id":42023348,"identity":"56c347c7-a4ef-4f1f-9d54-61000e895a39","added_by":"auto","created_at":"2023-08-23 16:07:39","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":503779,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3023846/v1/af109415-8d93-45a7-a11c-36639bcfe06c.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Who Leads? Who Follows? Exploring Agenda Setting by Media, Social Bots and Public in the Discussion of 2022 South Korea Presidential Election","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eSocial media has changed the traditional communication environment and agenda setting situation[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e],[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Social media provided an important platform to openly discuss top political and social issues[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e],[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e],[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. In contrast to traditional media, the gatekeeping power on social media has reduced [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e], content posted need not undergo editorial management nor scientific vetting [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e], [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], which have actually expanded the number and types of actors who potentially have the ability to dominate political discourse and shape agenda. So, the influence and shaping of public agendas are no longer the exclusive power of the news media in the age of social media[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Political parties, politicians, grassroots and etc are able to participate in shaping the agenda, which complicates the situation of agenda setting and creates the new questions of who sets political agendas in digital platforms like social media[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]?\u003c/p\u003e \u003cp\u003eEspecially, this complexity and uncertainty becomes more prominent as powerful emerging technologies evolve. Empirical studies have confirmed that social bots supported and manipulated by algorithms[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], have appeared in many political elections and events from more than ten countries around the world [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], including the 2016 US Presidential election and the 2017 French Presidential election[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Social bots seem to have become a mainstream tool of public opinion manipulation, indispensable in elections and other political activities. These automated accounts on social media, deployed by the actors behind them for a certain purpose, mimic the characteristics of human user and spread various information at a very high speed, including false information and junk news[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e],[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e],[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e],[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], which have affected the public opinion environment of social media. This has led us to wonder whether such social bots with powerful algorithm-based support can also participate in the shaping agenda? The question is the important starting point of this study.\u003c/p\u003e"},{"header":"2. Present research","content":"\u003cp\u003eAgenda-setting theory has been inseparable from political elections since its inception [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The study by McCombs and Shaw published in 1972 showed that there was a high correlation between agendas voters considered were important and agendas reported prominently in the media, namely, the mass media influence what people think about [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Given the increasing number of emerging media platforms, scholars have questioned the applicability of traditional media effect models, such as the agenda setting in this complex, fragmented media environment[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSome studies still suggest that while the digital age provides many opportunities for ordinary individuals and other organizations to spread information, their impact on public opinion is limited. On social media, traditional media, such as newspapers, would still publish online news by registering their official accounts. While other information on social media somewhat weakens media attention, media still plays a critical role in agenda setting [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. By analyzing tweets about Paris attack, researchers found that professional mass media organizations still hold a greater agenda-setting power than individual opinion leaders for setting the public agenda, as they obtain significantly more tweets, mentions, and replies from the public [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eHowever, some researchers think that agenda setting needs to be relooked into as the public agenda is taking a central stage through the new/social media in the digital and globalised era [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Meraz made a similar point after studying the blog and social media agenda setting. She believed the popularity of independent blog networks had allowed citizens more influence and power in setting news agendas[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. In addition, some actors with a significant resources, such as political parties, politicians have the opportunity to participate in the discussion and agenda setting on social media. A president like Donald Trump, who can used social media to lie to the public, had also agenda-setting effects [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFurthermore, social bots, algorithm-supported, automated tools deployed on social media, are also fighting for salience on political issues and attributes[\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e] (Vargo, 2018). The distinguishing feature of social bots is that they can systematically and repeatedly release large amounts of information in a short period of time and connect with more users in the social network[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. In general elections, social bots are often used to intervene in political discussions. For example, through orchestrating, to create the impression of an organic support for political actors with the help of social bots, or to systematically discredit the candidate and weaken his reputation [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. As the emerging political opinion climate maker, social bots not only disturbed the public opinion environment, but also brought new challenges to the traditional agenda setting situation[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOn March 10th, 2022, the 20th South Korea presidential election ended with Yoon Seok-yeol\u0026rsquo; victory margin of 0.73%, which is the narrowest margin of victory in any presidential election in South Korea. In fact, the 20th South Korea Presidential election has been unusually intense since the primary election. Several opinion polling since 2022 have shown that two main candidates, Lee Jae-myung and Yoon Seok-yeol, have received very close approval ratings, presenting an evenly matched situation. Under the situation, any party with significant resources would all probably deploy state of the art technologies to enact influence operations and other forms of manipulation of public opinion [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. If social bots were also deployed in debate of the South Korea presidential election, what impact would they have on the political discussion? But for present, this suppose is not determined.\u003c/p\u003e \u003cp\u003eA lot of research focuses on social media and political elections, with more research on predicting election results through social media data[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Or studying the impact of social media on the election, including the negative effects and prevention measures[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Some researchers also focus on the changes in the agenda setting on digital platforms, including the characteristics and changes of the media and other actors participating in the political agenda setting [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] ,[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e], [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. The research focusing on the participation of the social bots in the political election mostly focuses on the characteristics of the bots and the impact on the public opinion environment generated by bots[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e],[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e],[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eYet, we believe that social bots, which are greatly enhancing as the increasing sophistication of Artificial Intelligence[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], also have the potential to participate in agenda setting. And what would be the contact between the bot agenda, media agenda and public agenda in the complex ecosystem like social media? In traditional agenda-setting, the mass media determines what the public thinks about and how to think[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], that is, the media agenda takes precedence over the public agenda. So would the bot agenda also be influenced by the media agenda? And would the agenda by bots influence the public's perception of the relevant agenda? Therefore, this work takes the South Korea presidential elections as a case, based on determining the involvement of social bots in political discussions, to explore the contact between the media agenda, bot agenda and public agenda from the perspective of agenda setting. The specific research questions are as follows:\u003c/p\u003e \u003cp\u003eRQ1: What are the priorities of agenda perceived by media, bot and public?\u003c/p\u003e \u003cp\u003eRQ2: Is there correlations between the media agenda, bot agenda, and public agenda?\u003c/p\u003e \u003cp\u003eRQ3: Media agenda, bot agenda, and public agenda who takes the lead in time?\u003c/p\u003e"},{"header":"3. Methodology","content":"\u003cp\u003eThis study used the following data and methods to achieve our research objectives. First, tweets related to 2022 South Korea presidential elections were the material used for analysis. Second, we used Botometer to identify social bots, and divided all accounts into media accounts, social bots and human users. Third, to address the RQ1 of this study, cluster analysis would be used to cluster the tweets and the agendas included in the discussion of 2022 South Korea presidential elections would be determined. And the manual and computer-aided topic coding were used to determine the distribution of the agenda in media, bot and public accounts. Fourth, to solve the RQ2, we would analyze the correlation of media agenda, bot agenda, and public agenda using SPSS software. Fifth, to solve the RQ3, we used Almon Polynomial Lag to identify the time lag effect. Through the result of time lag, whose agenda was ahead in time would be judged. We introduce the way of data collection and the specific application of these methods in the following paragraphs.\u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Data Collection\u003c/h2\u003e \u003cp\u003eDiscussions related to 2022 South Korea presidential elections on Twitter were taken as the contents of the analysis in the study. Choosing Twitter as the data collection platform mainly considers the following reasons: (1) Twitter is currently the most popular tool adopted by news organization for content dissemination. (2) Twitter is one of the most-used social media by South Korea Internet users[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. (3) Currently, social bots are the most deployed on Twitter[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eData collection was conducted from February 15 to March 9, 2022. The two timestamps are very important time points in the South Korea Presidential elections. On February 15, the South Korea Presidential elections officially entered the election campaign stage, which means that the candidates will conduct a mass publicity and canvassing campaign. Voting for the Presidential elections officially ended on March 9. Meanwhile, given that this work targeted Twitter users from South Korea and their tweets, we only collected tweets in the language of Korean. Hashtags were used to search for related tweets. On Twitter, hashtags are often added \u0026ldquo; # \u0026rdquo; thus aid the formation specific themes and topics[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Before data collection, we identified a list of hashtags through the following steps.\u003c/p\u003e \u003cp\u003eFirst, we screened out the hashtags related to the South Korean presidential election from the top twitter trending hashtags between February 15 and March 9,2022. After that, combining the words in the headlines of major South Korea newspapers, we perfected the hashtags list. Last, we sent this hashtags list to five native Korean Twitter users aged between 20\u0026ndash;40 years for confirmation and manual supplementation[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. The six hashtags appearing in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e were finally confirmed and used. We collected all the tweets covered in these tags. The official API data interface service provided by Twitter was used to collect tweets. When collecting data, the privacy of Twitter users is respected without any personal information collected and displayed.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eHashtags were used to search for related tweets.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHashtags\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMeanings\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e#제20대대통령선거\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThe 2022 presidential election\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e#대선\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGeneral election\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e#윤석열\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYoon Seok-yeol\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e#국민의힘\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePeople Power Party\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e#이재명\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLee Jae-myung\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e#더불어민주당\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDemocratic Party of Korea\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Detection of Social Bots\u003c/h2\u003e \u003cp\u003eCurrently, there are many methods to identify social bots, including crowd sourcing-based recognition methods, identification methods based on social network information, and machine learning-based recognition methods [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. This work uses a machine learning-based recognition method, whose representative tool is Botometer, developed by Institute of Network Science of Indiana University and open to all users, which is a typical framework for machine recognition of social bots[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. The API calling method provided by Botometer pro was used to detect a large number of accounts. Botometer returns a bot score by extracting and analyzing more than 1,000 features, including users\u0026rsquo; tweeting frequency, behavioral characteristics, personal profiles, social networks, and friends\u0026rsquo; characteristics[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Scores closer to 1 represent a higher chance of being bots, while those closer to 0 are more likely to belong to humans[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Yet, at present, there is not agreement on a threshold that can reliably distinguish bots from humans. On the threshold setting, we want to refer to the complete automation probability (CPA) officially provided by Botometer and the studies had published. If an account has a higher score than the defined threshold, we will defined as a bot. It's also worth noting that Botometer fails to output a score when encountering account suspensions and authorization problems [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e], [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Text Cluster Analysis based on KH Coder Software\u003c/h2\u003e \u003cp\u003eIn the domain of natural language processing, clustering is an unsupervised process which is based on similarity between data to form some pattern [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. In this process, objects that are similar to each other are assigned a specific group or class from those that are dissimilar to them [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Various automation software and methods including K-Means, Affinity Propagation and others have been posed for clustering. For clustering analysis of texts, this work used KH Coder, an open source software for quantitative content analysis, text mining and computational linguistics[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Based on the clustering large applications (CLARA), a relatively large number of documents are able to process by KH Coder. Moreover, with the support of the CLARA algorithm, the KH Coder can realize the comparison of the sample clustering, and select the optimal center clustering [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eImportantly, multiple natural languages, including Korean, can be recognized and processed by KH Coder. Prior to performing the text clustering analysis, we first input a copy of Korean stop words into KH Coder, when processing it can automatically filters stop words out. After processing the tweets, KH Coder can return optimal clusters, each containing high probability words[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Based on the clustering results and the content of the original tweets, we would name each cluster returned by the KH coder, forming agendas of significant meaning. Later, each tweet was coded based on agendas by manual coding and computer-aided analysis, and the distribution of the agenda in media, bot and public accounts would be determined.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Correlation Analysis based on SPSS\u003c/h2\u003e \u003cp\u003eCorrelation analysis was tested for whether there was a significant relationship between the agendas of the two actors. We would analyze the correlation of media agenda, bot agenda, and public agenda using SPSS software. Media agenda, bot agenda, and public agenda were grouped pairwise and analyzed separately for correlation. The correlation of two types of agenda, such as the correlation between the media agenda and bot agenda, is judged by the hourly number of tweets posted under agendas. Firstly, the Kolmogorov-Smirnov test (KS text) was used to assess normality for data [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. KS test results showed that the data did not fit the normal distribution. So, Spearman correlation was used to determine the relationship. The significance level was set at p˂ 0.05 and the strong relationships set at p˂ 0.01 [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Time-Lag Analysis based on the Almon Polynomial Lag\u003c/h2\u003e \u003cp\u003eAlmon Polynomial Lag is a novel technique for estimating the weights of a distributed lag by means of a polynomial specification by Shirley\u0026middot;Almon[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. It transforms multiple jet-lag variables into a polynomial form that minimizes the multicollinearity problem and has the advantage of reducing the loss of free degree [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e] Since its introduction, the Almon lag technique has been widely used in empirical work. The main reason for its popularity is probably the ease with which it can be used-simply pick a length of lag, and a degree of polynomial, and results are quickly forthcoming [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. At present, the method is applied not only in econometrics and political economy, but also in social media data analysis[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Eviews software was used for analysis in this study and the analysis process as follows.\u003c/p\u003e \u003cp\u003eFirstly, determining the objects of the time-lag analysis. The objects of the time-lag analysis were determined based on the results of the correlation analysis. Two types of agendas with significant correlations would be the object of the time-lag analysis. Combinations without correlations were not objects of the time-lag analysis.\u003c/p\u003e \u003cp\u003eSecondly, determining the time units. The previous studies, used time series methods for agenda lag analysis, were set in years, month or days [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e],[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. However, in this study, we performed the analysis of time lags in hours. Two reasons are specifically considered. One is the time range of the data selected in our study is one month, the other is the information spreads fast on social media, especially, the number of topics and tweets would increase near the presidential election. The discrepancy in time between different agendas is more obvious by hours.\u003c/p\u003e \u003cp\u003eFinally, presenting the analysis results. After entering data normatively, the software could return the result of the lag order. The orders shown by results are the hours between the two agendas. For example, when x\u0026thinsp;=\u0026thinsp;bot, y\u0026thinsp;=\u0026thinsp;human, the result is fourth-order, which indicates that the bot agenda lags behind the public agenda by order 4, namely four hours.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Results","content":"\u003cp\u003eAs mentioned above, we obtained 45,169 tweets under six hashtags from February 15 to March 9,2022. The language of all the tweets is in Korean. To facilitate the subsequent data processing, repeated tweets, hyperlinks, and emoji symbols were filtered. After data cleaning, a total of 44,009 valid tweets posted by 18,700 users. According to the research approach of this work, after detecting social bots and dividing the three types of accounts, we conducted cluster analysis and time lag analysis successively, and obtained the corresponding results.\u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Social Bots Detection Results\u003c/h2\u003e \u003cp\u003eThrough a specific Python program, 18,700 Twitter users were entered to Botometer to obtain scores. The final scores of 17,362 accounts were returned, and 1,338 accounts could not be identified and detected by Botometer. So, the accounts and tweets that cannot be identified and detected by Botometer not be considered as the subjects of this study. 17,362 Twitter accounts and the 42,076 tweets they posted constitute the dataset of this study. We set the threshold to 0.8 to separate humans from bots. In other words, accounts with a score greater than 0.8 are bots, and accounts with less than 0.8 are human and media. Followed, we manually classified and counted the accounts of bots, humans and medias. There were 7,490 social bots, accounting for 43.14% of the total users in the dataset, who produced 18,206 tweets. 19,835 tweets were posted by 9,834 human accounts. And only 38 media accounts certified by Twitter, in the dataset, who published 4,035 tweets. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the number of the three types of accounts and their tweets.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Text Clustering Analysis and Agenda Distribution Results\u003c/h2\u003e \u003cp\u003eAfter processing the tweets, KH Coder returned 13 clusters. And 12 lists of words of practical meaning were marked with names, forming 12 agendas, which are displayed in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Because cluster 13 has no interpretable meaning, it was not presented. The 12 agendas can be divided into two categories. The one category is directly related to the South Korea general election, including economic resurgence, political liquidation, social welfare, Democratic Party of Korea, People Power Party and stigmatization propaganda. The other category is not directly related to the general election, including international relation, Russia-Ukraine conflict, gender issues, Seoul real estate, fasting dog meat and forest fire.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eList of 12 agendas.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgendas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh Probability Words\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInternational relation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e대한민국(South Korea), 미국(U.S.A), 중국(China), 일본(Japan), 북한에(Democratic People\u0026rsquo;s Republic of Korea), 캐나다(Canada), 호주(Australia), 터키(Turkey), 인도(India), 파키스탄(Pakistan), 러시아(Russia), 우크라이나(Ukraine)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRussia-Ukraine conflict\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e대한민국(South Korea), 러시아(Russia), 우크라이나(Ukraine),전쟁(war), 평화(peace), 관계를(relationship), 대통령이(president), 푸틴(Vladimir Putin), 젤렌스키(Zelensky), 나토는(the North Atlantic Treaty Organization), 충돌(conflict), 핵(nuclear weapon), 공격을( attack), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeol)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEconomic resurgence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e민생경제(livehood economy), 경제(economy), 회복을(recovery), 노동자(worker), 최저임금(minimum wage), 최저임금제도를(Minimum wage system), 알바(part-time job), 자영업자(individual household ), 중소상공인(middle and small industrialist and merchant), 부담을(burden), 부채(be in debt), 대출(loan),저소득(low-income)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePolitical liquidation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e대한민국(South Korea), 대선(general election), 대통령을 (president), 후보 (candidate),정권교체(regime change), 신천지(new field ), 정권(regime), 교체(replace), 전환(transition), 리셋 (reset),문재인(Moon Jae-in), 이재명(Lee Jae-myung), 더불어민주당(Democratic Party of Korea),국민의힘(People Power Party), 윤석열(Yoon Seok-yeol)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial welfare\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e복지국가(welfare state ), 복지(welfare), 노인(elder), 노인요양(The old man recuperation), 장애인돌봄(Care for the disabled), 건강보험(healtb-insurance), 휴게수당(recess allowance), 초등돌봄(Primary care), 아동(children), 수당(allowance), 보육(child welfare), 육아(child rearing),국가장학금(National scholarship), 대학생(college student)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender issues\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e남성(man), 여자(woman), 성질(sexual), 양성평등(sex equality),여성(gender), 성별(female), 주부(housewife),갈등(contradiction), 화합을(harmony), 혐오(disgust), 범죄(crime), 성범죄(sexual crime), 문제(problem),여성가족부를(Women's family department),페미니즘은(feminism),후보 (candidate)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSeoul Real Estate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e서울(Seoul), 부동산(real estate), 집값(housing price), 폭락(steep fall), 폭등(spurt in prices), 청년(youth), 내집마련(buy a house), 청년층(Youth order), 좌절감이(frustration), 불공정(unfair), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeo)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFasting dog meat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e대한민국(South Korea), 개고기(Dog meat), 유통시장(trading market), 모란시장(Peony market), 동물(animal), 반려동물 (pet), 법률(law), 금지(ban), 전통(tradition), 지원(support), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeol)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest Fire\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e대한민국(South Korea), 강원도(Gangwon-do),산불(wildfire), 삼림(forest), 산림( mountain forest), 화재(fire), 소방청(Fire hall), 소방대원(Firefighters), 이재민(victims of a natural calamity), 원자력(Nuclear), 발전소(Power Plant),현장(scene), 피해(lose), 대선(general election), 이재명(Lee Jae-myung), 윤석열(Yoon Seok-yeol)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDemocratic Party of Korea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e대한민국(South Korea), 대선(general election),더불어민주당(Democratic Party of Korea),후보가(candidate), 이재명(Lee Jae-myung),기호1번(No.1), 1번(No.1), 1번남(No.1 male),투표(vote), 문재인(Moon Jae-in), 지지율(support rate), 실현(support,) 나를위해(for me),외롭지(lonely)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePeople Power Party\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e대한민국(South Korea), 국민의힘(People Power Party), 후보가(candidate), 윤석열(Yoon Seok-yeol), 지지율(support rate), 2번남(No.2 male), 기호2번(No.2),투표(vote), 낙선이고(lose an election), 터무니없다(absurd)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStigmatization Propaganda\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e흑색선전(stigmatization propaganda),이재명(Lee Jae-myung), 개발사업(developmental project), 경기도(Gyeonggi), 대장동(),부동산(real estate), 아들(son), 성매매(Sex deals), 도박(gamble), 성매수(go whoring), 윤석열(Yoon Seok-yeol), 성접대(Sexual entertainment), 김건희(Kim Keon-hee), 주가조작(Manipulation of stock price), 통정매매(Overall trading), 거짓말(lie), 경력(qualifications), 학력(educational background), 위조에(counterfeit), 장모님(wife's mother)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThen, we analyzed the overall data with manual coding and Python program-assisted query, i.e., topic-coding for each tweet. When a tweet matches one or more keywords on an agenda, it is marked 1 by the program, and 0 when there is no match. Followed, we did the statistics for the number of tweets posted by three types of accounts under the 12 agendas after obtaining the encoding results to determine the distribution proportion of the agenda in media, bot and public accounts. Figure\u0026nbsp;2 presents the results.\u003c/p\u003e \u003cp\u003eIt can be seen that the media agenda did not involve in international relations. The bot agenda was more focused on economic resurgence, political liquidation, Democratic Party of Korea, People Power Party and stigmatization propaganda, and the public agenda was more focused on international relations, Russia-Ukraine conflict, gender issues, social welfare, Seoul real estate, fasting dog meat, and forest fire.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Correlation Analysis Results\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents the bivariate correlation coefficients under the 12 agendas and their significance results.The results show that the media agenda and bot agenda have significant correlation in social welfare, gender issues, Seoul real estate, fasting dog meat, Democratic Party of Korea, People Power Party, and stigmatization propaganda. Bot agenda and public agenda have significant correlation in the Russia-Ukraine conflict, economic recovery, political liquidation, social welfare, fasting dog meat, forest fire, Democratic Party of Korea and stigmatization propaganda. Media agenda and public agenda have a significant correlation in gender issues, economic resurgence, social welfare, Seoul real estate, fasting dog meat, and stigmatization propaganda.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eResults of the correlation analysis.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgendas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBivariate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003cp\u003eCoefficent\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003ep value\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eResults\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eInternational relation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.978\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eRussia-Ukraine conflict\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.116\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.068\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.172**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.006\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.087\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eEconomic resurgence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.361\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.242**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.192*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.032\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003ePolitical liquidation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.819\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.124**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.044\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.352\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eSocial welfare\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.147*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.315**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.299**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eGender issues\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.309**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.647\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.246**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eSeoul Real Estate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.386**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.682\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.354**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eFasting dog meat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.338**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.367**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.346**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eForest Fire\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.894\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.532**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.031\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eDemocratic Party of Korea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.271**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.176**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.075\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003ePeople Power Party\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.121**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.192\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.049\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUncorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eStigmatization propaganda\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Bot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.189**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBot-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.185**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedia-Human\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.105*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.026\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.5 Results of the Time-Lag Analysis\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e is the results of the time-lag analysis based on Almon polynomial. Overall, the time lags exist mainly between bot agenda and public agenda. Bot agenda lag behind public agenda in Russia-Ukraine conflict, fasting dog meat and forest fires. Public agenda lag behind bot agenda in economic resurgence, political liquidation, Democratic Party of Korea and stigmatization propaganda. Yet, the time lags does not exist between the media agenda and the other two types of agendas.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eResults of the time-lag analysis.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgendas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTime-lag (x\u0026thinsp;=\u0026thinsp;bot, y\u0026thinsp;=\u0026thinsp;human)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTime-lag (x\u0026thinsp;=\u0026thinsp;human, y\u0026thinsp;=\u0026thinsp;bot)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRussia-Ukraine conflict\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efourth-order\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEconomic resurgence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003esixth-order\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePolitical liquidation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003efirst-order\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFasting dog meat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eninth-order\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest fires\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003esixth-order\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDemocratic Party of Korea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ethird-order\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStigmatization propaganda\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e/\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ethird-order\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"5. Discussion","content":"\u003cp\u003eThrough exploring, we found that there were indeed social bots involved in election discussions on Twitter during the South Korea election. We would discuss and conclusion based on the results of this study.\u003c/p\u003e \u003cp\u003eFirstly, the priorities of the media agenda, bot agenda, and public agenda are different. On the agendas of social welfare, stigmatizing propaganda and the People Power Party, the proportion of media tweets is higher than on other agendas. Moreover, the media did not post tweets on social media related to the agenda of international relations. That is to say medias usually focused the agenda setting on domestic issues rather than diplomatic issues when faced the major domestic political events.\u003c/p\u003e \u003cp\u003eAlthough the bot agenda was also directly related to the presidential election, the content was more diverse compared with media. We also found that the bot agenda exhibited characteristics by regularity. They set the agenda by heavily forwarding existing news stories and messages on social media. So, they were active on some agendas, such as the agenda of stigmatization propaganda involved the candidates' past disgraceful events, and the agenda of political liquidation that has occurred in successive presidential elections. The public agenda involves more content than the media agenda and bot agenda. Unlike social bots, the public tends to express their viewpoint and opinions on general election-related agendas such as economic resurgence, political liquidation, and social welfare. And, they are very sensitive to hot issues and pay close attention to current events such as the Russia-Ukraine conflict and the forest fires in South Korea.\u003c/p\u003e \u003cp\u003eSecondly, we found that the agendas of the three type of actors are not isolated, the correlation results show that there are still significant relationships on part of the agenda between media and social bots, between media and the public, and between social bots and the public. For example, in the agenda of the fasting dog meat, there is a correlation between media and bots, between media and public, and between bots and the public. Although the issue is not directly related to the election itself, it is highly controversial in South Korea. Different attitudes toward \"eating dog meat\" had also become a way for candidates to gain support. It shows that controversial issue are something that media, bots and humans all pay attention to. And in the process of participating in the discussion, they have the situation of quoting and forwarding each other.\u003c/p\u003e \u003cp\u003eThirdly, the temporal order of the agendas of the three types of actors were presented through a time-lag analysis. The results showed that the media agenda was not ahead of the bot agenda and the public agenda in time, and that the time order only appeared between social bots and the public. That said, the media agenda does not lead the bot agenda and public agenda. We believe that the media agenda was not ahead of the public agenda and was related to the public distrust of the information disseminated by the media, especially on social media[\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. The reason why the media agenda is not ahead of the bot agenda may be that social bots tend to interact with the general public rather than with the media in this election discussion.\u003c/p\u003e \u003cp\u003eSocial bots lead the public on four agendas: economic resurgence, political liquidation, democratic Party of Korea, and stigmatization propaganda.In the face of these issues, social bots simply spread information through massive retweets, while humans published their own ideas and produced information based on specific issues. After spending a certain \"reaction time\" and \"buffer time\", people naturally lag behind bots, the automated tools for disseminating information based on algorithms.\u003c/p\u003e \u003cp\u003eThe public agendas lead bot agendas in Russia-Ukraine conflict, fasting dog meat and forest fires. The three agendas are in not directly related to the South Korea presidential election itself. But the people's discussions have still linked them to the presidential election. The forest fire in Gangwon-do during the election was seen by some superstitious people as a symbol of unlucky. As candidates Lee Jae-myung and Yoon Seok-yeol have different attitudes to fasting dog meat, people discussed the topic while expressing their support or not attitude for two candidates. Social bots operated mechanically that had not advantage on the discussion of these active agenda.\u003c/p\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eIn the digital age, the development of various emerging platforms and the use of new technologies complicate the communication environment. Many researchers believe that although the traditional news media was central in the agenda setting in the past, the agenda setting of the traditional media is is now just one force among many competing influences in the Internet environment [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMultiple actors, including politicians, the general public and technical tools represented by social bots, all fight for salience on political agendas and attributes [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. This work precisely focuses on and discusses the agenda-setting characteristics of media, social bots, and the public in the complex ecosystem of social media. Using a scientific approach, we first found the agenda used and avoided by three types of actors when facing the same event. While their agenda focus are different, they are relevant in some ways. On controversial agendas such as fasting dog meat, for example, there is a pairwise correlation between media, social bots and the public. On the basis of correlation, we further explore who takes the lead in time. We found that the media agenda was not ahead of the bot agenda and the public agenda in time. This suggests that the agenda setting monopoly power of the news media is somewhat weakened. Overall, our work reveals the dynamic features of multi-participant agenda settings on social media ecosystem where humans and bots to coexist, providing a viable path to parsing the complexity and diversity of agenda settings in digital environments.\u003c/p\u003e \u003cp\u003eOf course, there are also some limitations of this study. First, this work mainly focuses on the South Korea election as a case study, and whether the findings we reported generalize to other countries remains to be considered, which is also one of our future research directions. Second, although our results were obtained by a scientific analysis of the data collected, our data volume is relatively small. At the same time, we will explore better ways and technologies to obtain data in the future without violating the rules of the platform.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable. \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMenghan Zhang is responsible for researching architectural ideas, proposing research hypotheses, and analyzing the content. Ze Chen is responsible for collecting and analyzing preliminary data. Xinyan Liu is responsible for language proofreading of the research content. Jun Liu is responsible for implementing and liaising with the research, including coordination and content creation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study did not receive any funding in any form.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analysed during the current study are not publicly available due to ethical issues but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis manuscript is not under review elsewhere and the results have not been published previously or accepted for publication. This manuscript has been seen and approved by all authors. All methods were performed in accordance with the relevant guidelines and regulations. The questionnaire and methodology for this study was approved by the research ethics committee of the Soochow University and University of Copenhagen before data collection.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInformed consent was obtained from all participants included in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eYang, X., Chen, B.-C., Maity, M., Ferrara, E.: Social politics: agenda setting and political communication on social media. In: International Conference on Social Informatics, pp. 330\u0026ndash;344. Springer (2016)\u003c/li\u003e\n\u003cli\u003eGilardi, F., Gessler, T., Kubli, M., \u0026amp; M\u0026uuml;ller, S. (2022). Social media and political agenda setting. Political Communication, 39(1), 39-60.\u003c/li\u003e\n\u003cli\u003eCarlisle, J.E., Patton, R.C.: Is social media changing how we understand political engagement? an analysis of facebook and the 2008 presidential election. Polit. Res. Q. 66(4), 883\u0026ndash;895 (2013)\u003c/li\u003e\n\u003cli\u003eVarol, O., Ferrara, E., Ogan, C.L., Menczer, F., Flammini, A.: Evolution of online user behavior during a social upheaval. In: Proceedings 2014 ACM Conference on Web Science, pp. 81\u0026ndash;90 (2014)\u003c/li\u003e\n\u003cli\u003eYang, X., Chen, B.-C., Maity, M., Ferrara, E.: Social politics: agenda setting and political communication on social media. In: International Conference on Social Informatics, pp. 330\u0026ndash;344. Springer (2016)\u003c/li\u003e\n\u003cli\u003eCharles‐Smith, L.E.; Reynolds, T.L.; Cameron, M.A.; Conway, M.; Lau, E.H.Y.; Olsen, J.M.; Pavlin, J.A.; Shigematsu, M.; Strei‐chert, L.; Suda, K.J.; et al. Using Social Media for Actionable Disease Surveillance and Outbreak Management: A Systematic Literature Review. PLoS ONE 2015, 10, e0139701. https://doi.org/10.1371/journal.pone.0139701.\u003c/li\u003e\n\u003cli\u003eMeleo‐Erwin, Z.; Basch, C.; MacLean, S.A.; Scheibner, C.; Cadorett, V. \u0026ldquo;To each his own\u0026rdquo;: Discussions of vaccine decision‐making in top parenting blogs. Hum. Vaccines Immunother. 2017, 13, 1895\u0026ndash;1901. https://doi.org/10.1080/21645515.2017.1321182.\u003c/li\u003e\n\u003cli\u003eHindman, D.B., 2009. Mass media flow and differential distribution of politically disputed beliefs: the belief gap hypothesis. Journal. Mass Commun. Q. 86 (4), 790\u0026ndash;808.\u003c/li\u003e\n\u003cli\u003eFerrara, E.; Varol, O.; Davis, C.; Menczer, F.; Flammini, A. The rise of social bots. Commun. ACM 2016, 59, 96\u0026ndash;104.https://doi.org/10.1145/2818717.\u003c/li\u003e\n\u003cli\u003eWoolley S C, Howard P N. Political communication, computational propaganda, and autonomous agents: Introduction[J]. International Journal of Communication, 2016, 10.\u003c/li\u003e\n\u003cli\u003eFerrara, E. (2020). Bots, elections, and social media: a brief overview. Disinformation, Misinformation, and Fake News in Social Media, 95-114.\u003c/li\u003e\n\u003cli\u003eKim, A. Nearly Half of the Twitter Accounts Discussing \u0026lsquo;Reopening America\u0026rsquo; May Be Bots, Researchers Say. CNN. 22 May 2020. Available online:\u003c/li\u003e\n\u003cli\u003eShao, C.; Ciampaglia, G.L.; Varol, O.; Yang, K.‐C.; Flammini, A.; Menczer, F. The spread of low‐credibility content by socialbots. Nat. Commun. 2018, 9, 4787.\u003c/li\u003e\n\u003cli\u003eZhang, M., Chen, Z., Qi, X., \u0026amp; Liu, J. (2022). Could Social Bots\u0026rsquo; Sentiment Engagement Shape Humans\u0026rsquo; Sentiment on COVID-19 Vaccine Discussion on Twitter?. Sustainability, 14(9), 5566.\u003c/li\u003e\n\u003cli\u003eMcCombs, M. E., \u0026amp; Shaw, D. L. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36(2), 176\u0026ndash;187. https://doi.org/10.1086/267990.\u003c/li\u003e\n\u003cli\u003eGuo, L., \u0026amp; McCombs, M. (2011, May). Network agenda setting: A third level of media effects. In annual conference of the International Communication Association, Boston, MA.\u003c/li\u003e\n\u003cli\u003eYang, \u0026amp; Sun. (2021). Who has set whose agenda on social media? a dynamic social network analysis of tweets on paris attack. Communication Quarterly.\u003c/li\u003e\n\u003cli\u003eSalman, A., Mustaffa, N., Mohd Salleh, M. A., \u0026amp; Ali, M. N. S. (2016). Social media and agenda setting: Implications on political agenda. Jurnal Komunikasi Malaysian Journal Of Communication, 32(1), 401-414.\u003c/li\u003e\n\u003cli\u003eMeraz, S. (2009). Is there an elite hold? Traditional media to social media agenda setting influence in blog networks. Journal of computer-mediated communication, 14(3), 682-707.\u003c/li\u003e\n\u003cli\u003eSkewes, E. (2018). Time delays are not enough; Media must call out lies. Journal of Media Ethics, 33(2), 97\u0026ndash;99. https://doi.org/10.1080/23736992.2018.1435498\u003c/li\u003e\n\u003cli\u003eCameron, M. P., Barrett, P., \u0026amp; Stewardson, B. (2016). Can social media predict election results? Evidence from New Zealand. Journal of Political Marketing, 15(4), 416-432.\u003c/li\u003e\n\u003cli\u003eAral, S., \u0026amp; Eckles, D. (2019). Protecting elections from social media manipulation. Science, 365(6456), 858-861.\u003c/li\u003e\n\u003cli\u003eMetaxas, P. T., \u0026amp; Mustafaraj, E. (2012). Social media and the elections. Science, 338(6106), 472-473.\u003c/li\u003e\n\u003cli\u003eSalman, A., Nguyen, C., Lee, Y. H., \u0026amp; Cooksey-James, T. (2016). A review of barriers to minorities\u0026rsquo; participation in cancer clinical trials: implications for future cancer research. Journal of immigrant and minority health, 18(2), 447-453.\u003c/li\u003e\n\u003cli\u003eLanger, A. I., \u0026amp; Gruber, J. B. (2021). Political agenda setting in the hybrid media system: Why legacy media still matter a great deal. The International Journal of Press/Politics, 26(2), 313-340.\u003c/li\u003e\n\u003cli\u003eHagen L, Neely S , Keller T E , et al. Rise of the Machines? Examining the Influence of Social Bots on a Political Discussion Network[J]. Social Science Computer Review, 2020(10):089443932090819.\u003c/li\u003e\n\u003cli\u003eSch\u0026auml;fer, F., Evert, S., \u0026amp; Heinrich, P. (2017). Japan\u0026apos;s 2014 general election: Political bots, right-wing internet activism, and prime minister Shinzō Abe\u0026apos;s hidden nationalist agenda. Big data, 5(4), 294-309.\u003c/li\u003e\n\u003cli\u003ePastor-Galindo, J., Zago, M., Nespoli, P., Bernal, S. L., Celdr\u0026aacute;n, A. H., P\u0026eacute;rez, M. G., ... \u0026amp; M\u0026aacute;rmol, F. G. (2020). Spotting political social bots in Twitter: A use case of the 2019 Spanish general election. IEEE Transactions on Network and Service Management, 17(4), 2156-2170.\u003c/li\u003e\n\u003cli\u003ePublished by Nina Jobst, Apr 11, 2022 Social media usage in South Korea - Statistics \u0026amp; Facts.https://www.statista.com/topics/5274/social-media-usage-in-south-korea/\u003c/li\u003e\n\u003cli\u003eBruns, A., \u0026amp; Burgess, J. (2011). The use of Twitter hashtags in the formation of ad hoc publics. In Proceedings of the 6th European consortium for political research (ECPR) general conference 2011 (pp. 1-9). The European Consortium for Political Research (ECPR).\u003c/li\u003e\n\u003cli\u003eBadawy, A.; Lerman, K.; Ferrara, E. Who Falls for Online Political Manipulation? In Proceedings of the Companion Proceedingsof the 2019 World Wide Web Conference, Association for Computing Machinery, San Francisco, CA, USA, 13\u0026ndash;17 May 2019; pp.162\u0026ndash;168.\u003c/li\u003e\n\u003cli\u003ePing, H. , \u0026amp; Qin, S. . (2018). A Social Bots Detection Model Based on Deep Learning Algorithm. 2018 IEEE 18th International Conference on Communication Technology (ICCT). IEEE.\u003c/li\u003e\n\u003cli\u003eLuceri, L.; Deb, A.; Giordano, S.; Ferrara, E. Evolution of bot and human behavior during elections. First Monday 2019, 24, 1\u0026ndash;12.https://doi.org/10.5210/fm.v24i9.10213.\u003c/li\u003e\n\u003cli\u003eDavis, C.A.; Varol, O.; Ferrara, E.; Flammini, A.; Menczer, F. BotOrNot: A system to evaluate social bots. In Proceedings of theProceedings of the 25th International Conference Companion on World Wide Web, Montreal, QC, Canada, 11\u0026ndash;15 April 2016;pp. 273\u0026ndash;274. https://doi.org/10.1145/2872518.2889302.\u003c/li\u003e\n\u003cli\u003eFerrara, E. Measuring Social Spam and the Effect of Bots on Information Diffusion in Social Media. In Complex Spreading Phe‐nomena in Social Systems; Springer: Berlin/Heidelberg, Germany, 2018; pp. 229\u0026ndash;255. https://doi.org/10.1007/978‐3‐319‐77332‐2_13.\u003c/li\u003e\n\u003cli\u003eAhuja, S. , \u0026amp; Dubey, G. . (2017). Clustering and sentiment analysis on Twitter data. 2017 2nd International Conference on Telecommunication and Networks (TEL-NET). IEEE. \u003c/li\u003e\n\u003cli\u003eJ. C. Dunn, \u0026ldquo;A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters,\u0026rdquo; Journal of Cybernetics 1973, 3: 32-57.1080/13501763.2018.1458890\u003c/li\u003e\n\u003cli\u003eHiguchi, K. (2016). KH Coder 3 reference manual. Kioto (Japan): Ritsumeikan University.\u003c/li\u003e\n\u003cli\u003eHinton, P., McMurray, I., \u0026amp; Brownlow, C. (2014). SPSS explained. Routledge.\u003c/li\u003e\n\u003cli\u003eAlmon, S. (1965). The distributed lag between capital appropriations and expenditures. Econometrica: Journal of the Econometric Society, 178-196.\u003c/li\u003e\n\u003cli\u003ePeter Schmidt \u0026amp; Roger N. Waud (1973) The Almon Lag Technique and the Monetary versus Fiscal Policy Debate, Journal of the American Statistical Association, 68:341, 11-19\u003c/li\u003e\n\u003cli\u003eLehrer, S., Xie, T., \u0026amp; Zeng, T. (2021). Does high-frequency social media data improve forecasts of low-frequency consumer confidence measures?. Journal of Financial Econometrics, 19(5), 910-933.\u003c/li\u003e\n\u003cli\u003eDearing, James W., and Everett M. Rogers (1988) \u0026quot;The agenda-setting process for the issue of AIDS.\u0026quot; Paper presented to the Mass Communication Division, International Communication Association, New Orleans, 29 May-3 June.\u003c/li\u003e\n\u003cli\u003eSmith, Kim A. (1987) \u0026quot;Newspaper coverage and public concern about community issues: A time-series analysis.\u0026quot; Journalism Monographs 101:1-32. Smith, Ted J., Ill, and J. Michael Hogan (1987)\u003c/li\u003e\n\u003cli\u003eSe-Uk Oh So-Eun Lee, 2021 Digital News Report https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2021/south-korea\u003c/li\u003e\n\u003cli\u003eVargo, C. J. (2018). Fifty years of agenda-setting research: New directions and challenges for the theory. The Agenda Setting Journal, 2(2), 105-123.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"agenda setting, social media, social bots, Twitter, political communication, general election","lastPublishedDoi":"10.21203/rs.3.rs-3023846/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3023846/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSocial media not only changes the traditional communication environment, but also brings new changes to agenda setting. The main body of agenda setting has shifted from the traditional media to the politicians, political parties and grassroots people. With the increasing use of social bots in public opinion manipulation and political election interference, whether they can participate in or influence agenda setting has become an urgent concern. So far, there is less literature focusing on engagement in agenda-setting for social bots. This paper studies the social media discussion content of the South Korean presidential election, determines the participation of social bots, and explores the connection between media agenda, bot agenda and public agenda from the perspective of agenda setting. The study found that while the main agendas of media, social bots and the public are not the same, their agendas are relevant. In addition, the media agenda is not timely ahead of the bot agenda and the public agenda, and the time order only appears between the social bots and the public.\u003c/p\u003e","manuscriptTitle":"Who Leads? Who Follows? Exploring Agenda Setting by Media, Social Bots and Public in the Discussion of 2022 South Korea Presidential Election","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2023-06-09 15:57:10","doi":"10.21203/rs.3.rs-3023846/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7ce5400a-f2a7-456d-a138-dd9a581c338f","owner":[],"postedDate":"June 9th, 2023","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2023-08-23T15:59:32+00:00","versionOfRecord":[],"versionCreatedAt":"2023-06-09 15:57:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3023846","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3023846","identity":"rs-3023846","version":["v1"]},"buildId":"-HB7Z8yhvgn0wM9Nzuekk","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}