The emerging viewertariat in South Korea: The Seoul mayoral TV debate on Twitter, Facebook, and blogs

Telematics and Informatics 33 (2016) 570–583

Contents lists available at ScienceDirect

Telematics and Informatics journal homepage: www.elsevier.com/locate/tele

The emerging viewertariat in South Korea: The Seoul mayoral TV debate on Twitter, Facebook, and blogs Yun-Cheol Heo a,b, Ji-Young Park c, Ji-Young Kim d, Han-Woo Park d,⇑ a

Dept. of Communication, Pusan National University, South Korea Cyber Emotions Research Center, YeungNam University, South Korea Dept. of East Asian Cultural Studies, YeungNam University, South Korea d Dept. of Media & Communication, YeungNam University, South Korea b c

a r t i c l e

i n f o

Article history: Received 17 August 2014 Received in revised form 3 May 2015 Accepted 5 August 2015 Available online 11 August 2015 Keywords: Viewertariat South Korea Election Social media Twitter Facebook Blog TV debate

a b s t r a c t Social networking sites (SNSs) represent Web 2.0 platforms or networking tools through which users can freely exchange ideas, opinions, experiences, and viewpoints and thus have considerable influence on the formation of political discourse. Despite the wide diffusion of SNSs and their increasing political influence, traditional media such as TV have retained their influence to a certain extent because new media and traditional media are not independent of each other. In particular, recent technological advances have made it possible for individuals to exchange their opinions through SNSs on a real-time basis while watching TV. As a result, the formation of political discourse may shift from the traditional mass media to social media, and viewer responses generated through social media may be transferred quickly to the mass media. Given this important trend, this study provides an empirical analysis of the pattern of interactions between TV and SNSs in the Korean context. More specifically, the study investigates the features and patterns of online messages from SNS users in Korea about TV debates during the Seoul mayoral by-election in 2011. By assuming some differences in features of political discourse across various types of SNSs, the study compares those features specifically associated with TV debates by considering Twitter, Facebook, and blogs. The results suggest that SNS users not only accept and interpret political discourse while watching TV but also participate actively in its production and restructuring. In addition, the results indicate some differences in communication patterns between the three SNS platforms. Ó 2015 Elsevier Ltd. All rights reserved.

One of the most important changes in journalism in recent years has been the worldwide diffusion and popularization of social networking sites (SNSs). SNSs represent open-media platforms or networking tools that allow users to freely exchange ideas, opinions, experiences, and viewpoints and thus have considerable influence on the formation of political communication and discourse. According to typical explanations of the formation of political discourse through the mass media, various media can highlight the importance of specific issues (McCombs and Shaw, 1972). Previous studies have suggested that elite segments of society tend to be the main definers of important social events and thus that they define the significance of such events in a passive manner (Hall, 1982). However, the emergence of SNSs, which can form hubs that expand interactive

⇑ Corresponding author. E-mail address: [email protected] (H.-W. Park). http://dx.doi.org/10.1016/j.tele.2015.08.003 0736-5853/Ó 2015 Elsevier Ltd. All rights reserved.

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

571

communication while sharing and amplifying citizens’ political interests, requires a new approach to understanding the formation of political discourse through the mass media. With the rise of diverse forms of media that connect individuals to one another, media have been understood not only as channels for distributing messages but also as a space in which individuals can self-organize their opinions on specific issues (Shirky, 2008). Castells (2009) explained the changes in power relationships caused by recent developments in the media environment by suggesting the concept of mass self-communication and emphasizing networking power, which refers to an individual’s power over who or what is included in a network. Here mass self-communication refers to the use of new media for private and/or public messages that can reach the masses. In particular, recent technological advances have made it possible for individuals to exchange their opinions through SNSs on a real-time basis while watching TV. As a result, the formation of political discourse may shift from the traditional mass media to social media, and viewer responses generated through social media may be transferred quickly to the mass media. Given this important trend, this study empirically analyzes the pattern of interactions between TV and SNSs in the Korean context. More specifically, the study investigates the features and patterns of online messages from SNS users in Korea about TV debates during the Seoul mayoral by-election in 2011. By assuming some differences in features of political discourse across various types of SNSs, the study compares those features specifically associated with TV debates by considering Twitter, Facebook, and blogs. 1. A literature review 1.1. Emergence of the viewertariat Television is a centralized medium for top-down editorial control, whereas SNSs can facilitate horizontal connections between users. In addition, SNSs can provide TV viewers with many opportunities to communicate with one another on a real-time basis while watching TV. According to the Pew Research Center’s Internet & American Life Project (Smith and Boyles, 2012), 52% of adult users of mobile phones in the U.S. used their phones while watching TV. In particular, 29% made recent use of their phones to exchange text messages with other users who were watching the same TV programs but were in different locations. This trend suggests the emergence of a new group of viewers who maintain their individual identity and behave individually in ordinary times but behave collectively when specific issues of common interest arise (Hardt and Negri, 2004). This study considers the viewertariat, an interesting concept referring to the increasing trend toward digitally mediated social networking and the sharing of TV content. According to Anstead and O’Loughlin’s (2009) definition, this concept suggests a change in the passive role of TV viewers from a traditional perspective. The word ‘‘viewertariat” combines the words ‘‘viewer” and ‘‘proletariat” to indicate individuals’ exchange of opinions through SNSs while watching TV programs reflecting some political discourse. Anstead and O’Loughlin, 2009 coined this term in a study for Question Time, a popular TV program by BBC (this study was aired on October 22, 2009). They proposed this concept to refer to those viewers who exchanged opinions and debated through Twitter while watching Nick Griffin, the leader of the ultraright British National Party, on TV. Wohn and Na (2011) analyzed Tweets of those Twitter users who were posting those Tweets while watching President Barack Obama’s live speech at the White House announcing his acceptance of the Nobel Peace Prize on October 9, 2009. Ampofo et al. (2011) recently applied this concept to a TV debate during the British general election in 2010 and produced similar findings. This general election (the most recent one in the U.K.) illustrates how TV and SNSs can complement each other and coevolve (Ampofo et al., 2011; Anstead and O’Loughlin, 2011; Chadwick, 2011a,b). The U.K. has relatively advanced party politics, and therefore TV debates have not been a widely accepted part of elections because such debates have generally been viewed as tools for building popularity, not as platforms for discussing important policies. However, in 2010, the ruling and opposition party candidates agreed to hold the first TV debate during the general election, receiving considerable attention from the mass media and the public. In addition, SNSs had considerable influence on this general election. For example, the British Conservative Party posted Tweets about their campaign donations on MyConservatives.com to help SNS users to make convenient donations through Paypal. The Labor Party filled the first page of Labor Doorstep with supporters’ comments collected from Twitter and Facebook. In addition, the Liberal Democratic Party, led by Nick Clegg, who caused a sensation in the 2010 general election and was given the nickname ‘‘Nick Obama,” lampooned rival parties through Labservative.com. The Liberal Democratic Party had the most visible online presence and had the largest number of Twitter followers (Twitter.com/nick_clegg) (CNN, 4 May, 2010; The Guardian, 11 April, 2010). The key aspect of this election was the introduction of diverse services combining TV and SNSs. ITV showed viewers’ Tweets on their on-air page on a real-time basis by using the services of CoveritLive, a Canadian firm. Tweetminster (http://tweetminster.co.uk), a political website, introduced crude Tweets by viewers of the TV debate on a real-time basis by using the ‘‘sentiment tracker,” which analyzes the content of Tweets. In addition, BBC, Sky, and ITV transferred real-time responses from small audience panels through a real-time worm chart (Chadwick, 2011a). That is, during this election, the formation of political discourse shifted from the mass media to social media, and viewer responses generated through social media were then quickly transferred to the mass media. Based on this election, Chadwick (2011a) reported that the news cycle, which has traditionally been produced and distributed by an elite group of journalists, political parties, and power bloggers, is being replaced with the so-called ‘‘political information cycle,” which is formed collectively by individuals’ actions and instant responses through the fusion of traditional media and SNSs.

572

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

Based on the above discussion, the viewertariat may be a core concept with important implications for the social meaning of newly emerging groups of TV viewers from the perspective of the rapid diffusion of SNSs. Freelon and Karpf (2015) analyzed the viewertariat based on social media comments during the 2012 U.S. presidential debates and found the emergence of nontraditional political actors as prominent network hubs. In sum, the characteristics of the viewertariat are distinct from those of traditional opinion leaders (Weimann et al., 2007). First, unlike the latter, the former shows real-time responses to specific events. The traditional mass media require some time to distribute comments from opinion leaders. Second, conventional opinion leaders generally show a high level of education as well as socioeconomic status, whereas the viewertariat tends to be heterogeneous. Third, because of the distinct features of SNSs (i.e., many-to-many communication), the boundary between the viewertariat that influences other TV viewers and the viewertariat that is influenced by them is neither clear nor absolute. This is different from opinion leaders suggested by a two- or multi-step flow model in media and communication science. Fourth, the viewertariat is an ‘‘interpretation” group as well as an ‘‘action” group because it participates directly in issues while actively interpreting those issues provided by the mass media. This means that the viewertariat may serve as a critical mass, not as a depersonalized and anonymous mass. 1.2. Understanding the viewertariat through various SNSs Early television studies of uses and gratification theory indicated that people share their feelings, exchange their opinions with others, maximize their pleasure, and seek to satisfy their needs during viewing (Rubin, 1981). This tendency is also persistent on SNSs. According to a study in Belgium, Tweets generated during current affairs television TV programs had many discussions reflecting sarcasm, playfulness, and fun (D’heer and Verdegem, 2015). In addition, the adoption and widespread dissemination of SNSs across viewers do not reduce TV viewing but increases it through the vitalization of back-channel communication (Proulx and Shepatin, 2012; Stefanone et al., 2010). Hwang and Lim (2015) surveyed Korean college students during the 2012 London Olympic Games in terms of their social viewing experiences (e.g., commenting on SNSs and viewing TV programs at the same time) and found that the stronger the social viewing behavior, the stronger the commitment to the sports channel. However, care must be taken in interpreting this to mean the continued effect of social viewing regardless of the context. In the context of IT use, adaptive structuration theory (DeSanctis and Poole, 1994; Chin et al., 1997; Larsson, 2012a,b) posits that internet users are influenced and restricted by certain structures that can be changed through consistent communication and social interactions between these users. The interactive relationship between actors and structures in which there is mutual adaption is referred to as the adaptive structuration process. Therefore, the platform structure may influence the pattern of interactions between TV viewers in terms of their simultaneous use of SNSs. In this regard, various SNSs such as Twitter, Facebook, and blogs, which generally have distinct technological features (Table 1), may be structuralized in unique ways according to the features and norms of their users. Twitter is a type of microblog that allows the exchange of simple sentences of up to 140 words. Its simplicity allows Twitter users to exchange text messages, photos, and video/audio clips with their followers on a real-time basis (Java et al., 2007; Zhao and Rosson, 2009). The most unique feature of Twitter is its retweet function, which enables the forwarding of Tweets and thus the wide diffusion of messages in a short period of time. Twitter has an open-relationship platform in which no permission from other users is necessary for starting relationships, and therefore it facilitates weak ties. Because of these technological features, previous studies have often highlighted the similarity between Twitter and TV in terms of their one-to-many mode of communication (Barash and Golder, 2010). The mode of information sharing on Twitter differs from that on Facebook, which is an information-sharing platform based on networks of friends, not on those of general users. If Twitter is a market that facilitates the online exchange of information, then Facebook is a living room where friends share their life experiences through the publication of ‘‘status updates” (Page, 2012). Accordingly, Facebook can be a channel for information sharing and more credible than Twitter, but instant communication is less powerful on Facebook because its users tend to be pressured in terms of posting trustworthy messages of expert quality (Shih, 2009). Twitter and Facebook are well known for their ability to quickly attract the public’s attention and amplify information through mobile media. On the other hand, blogs are generally produced using desktop computers because bloggers typically post lengthy and reflective articles with photos and figures. Blogs, originally designed as brief internet diaries for recording bloggers’ thoughts (Blood, 2005), now facilitate active communication between bloggers and help to recreate communal content, facilitating the formation of the blogosphere, which is composed of networks of bloggers with some common interests (Bruns, 2008).

Table 1 Technological features of Twitter, Facebook, and blog. Type

Blog

Facebook

Twitter

Synchronous Word length Information attribution User relationship Korean users

Low No limit Professional Symmetric or asymmetric 8.0 million

Medium 5000 words Multifarious Symmetric 5.3 million

High 140 words Simplistic Asymmetric 5.4 million

Source: Compiled by the authors based on Hansen et al. (2010).

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

573

According to the Korea Internet and Security Agency (2012), 67.1% of all internet users used SNSs (male: 68.7%; female: 65.3%). People in their twenties were most likely to use SNSs (90.2%), followed by those in their teens (79.8%), thirties (72.2%), and forties (52.3%). In addition, 81.7% used blogs; 21.1%, Facebook; and 13.9%, Twitter (Korea Internet and Security Agency, 2012). As shown in Table 1, Korea had two distinct characteristics: bloggers and young individuals (those in their twenties) were most likely to use SNSs. However, the Korea Internet and Security Agency (2012) did not examine the unique number of users for each type of media. As discussed earlier, these technological features may determine the communication mode or culture of users because the unique features of each SNS can be structuralized differently according to the specific context. According to adaptive structuration theory, which is based on Giddens’s (1984) structuration theory, a user of a specific technology is affected by the social structure. Conversely, users may change the structure through their continued use of the technology and mutual interactions. Duality of media theory (derived from structuration theory) argues that some characteristics of media platforms are outcomes of interactions between users, structures, and mutual influence (Larsson, 2012; Webster, 2011; Anstead and O’Loughlin, 2015). This suggests a need for a comparison of various SNS platforms instead of investigating a single SNS. However, previous studies of the viewertariat are limited in that they have considered only one SNS, namely Twitter. Pressgrove (2012) highlighted different roles of Twitter, Facebook, and blogs in information diffusion but did not analyze messages posted by Twitter users, Facebook users, and bloggers who post their messages while watching TV. Therefore, this study focuses mainly on the technological differences between Twitter, Facebook, and blogs, their structuralization in the Korean context, and their interactions with TV. 1.3. Korean SNSs and the Seoul mayoral by-election The development of new point-to-point communication technologies such as SNSs facilitates democracy across the world by circulating political information and enabling horizontal communication (Skoric and Park, 2014). Korea is known for having a horizontal political culture but has witnessed some changes in horizontal communication regardless of the political ideology, that is, conservatives and progressives. Therefore, it may be problematic to draw some political implications of SNS use only from an increase in the number of SNS users and SNS use. Both the traditional mass media and new communications technologies had mixed effects on the 2012 Seoul by-election. In addition, because different types of SNSs continue to appear and are diffused, the emergence of a new type of structuration is likely. In this regard, the present study compares the use of Twitter, Facebook, and blogs by the viewertariat by considering the by-election for the mayor of Seoul on October 26, 2011. Korea is one of the world’s most connected countries in terms of the penetration of the high-speed wireless Internet (Reuters, 2011) and thus can offer some interesting and meaningful insights into the relationship between SNSs and politics. The recent decade has witnessed the rise of the Internet in Korea as a key factor influencing political changes and elections in the country. For example, previous studies have noted the political influence of SNSs on election outcomes after the completion of the nationwide mayoral elections in 2010 and the reelections in 2011 (Cho et al., 2011; Hsu and Park, 2011; Lim and Park, 2011; Sams et al., 2011). In this regard, with the resignation of Se-Hoon Oh as the mayor of Seoul, the mayoral by-election in 2011 became a major topic of discussion in terms of the role of SNSs. The Seoul mayoral by-election was held when Oh, a member of the ruling and conservative Grand National Party (GNP), which is comparable to the Republican Party in the U.S., resigned abruptly. Oh’s resignation resulted from a low voter turnout for the Seoul Free Lunch Referendum. The by-election was seen as a symbolic referendum on the Myung-Bak Lee administration’s performance in the last four years. In addition, it was considered a barometer for revealing the winning party in the 19th National Assembly election in April 2012 as well as in the presidential election in December 2012. However, the leading opposition party was not able to field a candidate for the by-election and thus had to support Won-Sun Park, a lawyer-civic activist, as a candidate representing unified opposition. There were two major candidates in the by-election: Kyoung-Won Na, who was well known for being attractive and gained some popularity among the general public because of her role as a GNP spokesperson, and Park, who was relatively new to the public. This raises the question of how Park won the by-election. There may be several reasons, but an investigation of domestic politics is beyond the scope of this study. One distinct characteristic of the by-election was that Na did not live up to her supporters’ expectations during the TV debate but that Park was evaluated to leverage the power of TV and social media to publicize his policy agendas and mobilize supporters as well as a majority of undecided voters (Lee and Park, 2012). Televised debates by political candidates were first introduced during the Seoul mayoral election in 1995 and have since been used in presidential, general, and local elections. The Public Official Election Act mandates the Election Broadcasting Debate Commission to organize debates during these elections (Election Broadcasting Debate Commission, 2010). In Korea, TV debates have generally been understood as a tool for providing voters with good opportunities for evaluating the competence and qualification of candidates and been assumed to be an important factor influencing election outcomes. Kim and Kim (2004) examined Korean voters’ media use during the 17th general election in 2004 and found that they evaluated TV as the most reliable and influential communication channel, despite the gradual decline in the role of the mass media. In terms of their political influence, traditional media such as TV still remain a powerful force because, as discussed earlier, new media and the mass media are not independent of each other. According to a recent report by the Korea National Statistics Office, the number of Twitter users in Korea jumped from 630,000 in June 2010 to 5,440,000 in December 2011, representing a 863.5% increase, and that of Facebook users increased from 4,010,000 in September 2011 to 5,360,000 in

574

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

December 2011, representing a 33.7% increase. Tweetmix.net, a provider of Twitter data in Korea, reported that the reelection in April 2011 generated approximately 100,000 Tweets about the election and candidates but that the Seoul mayoral by-election, which took place only six months later, generated approximately one million Tweets. This is a huge number, particularly in the context of Korea’s population (approximately 48.87 million as of 2011). Based on these trends in Korea, Shirky (2011) suggested the potential emergence of a new form of democracy mediated by social media in which SNSs represent an influential communication tool, explaining that internet users are increasingly showing their intention to participate in politics. Chadwick (2011a, 2011b) proposed a new political information cycle produced, organized, and disseminated by a large number of nonelite individuals, e.g., the viewertariat. This cycle is different from the conventional news cycle, which is facilitated through a small number of elites. In this cyclical process of political news, however, SNSs do not necessarily replace TV but redefine its political influence. Lin (2011) reviewed the literature on media substitution and complementarity and concluded that old and new media compete with each other but do not always nullify the other during the adoption of new media and their continued use. Kim (2010) suggested an agenda-weaving model based on the idea that, in terms of establishing and mediating agendas, traditional media and social media in Korea have not a hostile relationship but a cooperative one. Previous studies of news cycles in Korea’s online environment have focused mainly on explaining the close interaction between professional reporters and bloggers through case studies. For example, Park and colleagues (Chang and Park, 2012; Park and Jankowski, 2008; Park et al., 2011) found that the voice of individual citizens can be amplified in the news blogosphere and can politicize SNSs when they identify and modify hidden content on ordinary websites and disseminate it through cyberspace. In other words, recent trends in Korea suggest that anyone can become an opinion leader if he or she can produce information that can draw sympathy and attention in the blogosphere. In addition, Cho et al. (2011) considered local elections in June before the Seoul mayoral by-election in 2011 and reported that active individuals in the blogosphere are likely to respond sensitively to political agendas and make efforts to set and form public opinion by using newer social media such as Twitter and Facebook. In this regard, the present study focuses on the use of blogs, an early form of SNSs. In particular, the study examines the major features of messages posted by viewers of the Seoul mayoral TV debate by platform (Twitter, Facebook, and blogs) and analyzes the differences in semantic characteristics of those messages. 2. Research question SNSs have played an important role in fostering political interest and motivating people to vote (Bond et al., 2012; Han, 2012). The emergence of a viewertariat with some mixed use of both TV and SNSs has called attention to the integration of political discourse and dialogue. TV has provided people with conversation topics and discussion issues by serving as a vital cultural forum (Lull, 1990; Newcomb, 1994), and it has become easier for people to share their feelings and opinions on TV programs within online media environments (Bondad-Brown et al., 2012). In particular, the global diffusion of social media has given rise to social viewing, allowing social media to function as an integral outlet (Hill and Benton, 2012). Political and electoral information competes in diverse forms and is reorganized, reprocessed, and reproduced by heterogeneous individuals belonging to nonelite groups. During this process, old as well as new media undergo rapid hybridization and differentiation. Here the adaptive structuration of media platforms finally occurs through interactions between technological and social factors (DeSanctis and Poole, 1994). In particular, Leavitt et al. (2009) explained that for platforms developed after Web 2.0, diverse patterns of interactions can be expected because users can access media in diverse ways depending on their objectives. The present study investigates the features of the political discourse on the Seoul mayoral TV debate on various SNS platforms. In this regard, this study is guided by the following research questions: (RQ1) How do SNS users express their opinions during and after a TV debate? (RQ2) What are the characteristics of opinions in messages across SNS media in terms of the use of keywords? 3. Methods 3.1. Webometrics and the semantic network analysis In this study, the webometric technique, a set of research methods for analyzing various sociocultural aspects of individuals’ web-mediated communication activities involving particular issues, was used (Park, 2010). For example, some internet research methods such as hyperlink analysis, semantic network analysis, online survey, focused interview, search engine evaluation, virtual ethnography, and network visualization methods are often used in combination. In other words, webometrics is interdisciplinary in nature and can be broadly defined as a mixed method for analyzing web-based content (e.g., text, images, audiovisual objects, and hyperlinks) based mainly quantitative metrics (Ackland, 2013; Thelwall, 2009). Ackland (2013) emphasized that webometrics has become an active research domain with the development of methodologically sophisticated concepts, large data sets, and internet-based research tools for social scientists. However, Lim and Park (2011, 2013) noted that social scientists employing webometrics have focused mainly on analyzing hyperlink data and thus neglected web mentions of specific words, issues, and actors. To address this gap in the literature, some semantic differences between individualized Web 2.0 platforms were examined using rich descriptive data from postings. Few studies have provided a comparative analysis of various SNS platforms. In addition, few studies have employed the

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

575

webometric technique to analyze the interaction between traditional media and the webosphere. In this regard, an in-depth semantic network analysis of webmentions collected from three SNS platforms (Twitter, Facebook, and blogs) was conducted using visualization techniques based on information science and social network analysis. In addition, for a better understanding of SNS-mediated TV-viewing activities and information-sharing behaviors by platform, viewers’ responses were examined by reading their posts. Further, some focused interviews were conducted with individuals who frequently posted comments during the study period and actively engaged in discussions on the Seoul mayoral TV debate. This methodological strategy was employed in conjunction with the visualization and qualitative analysis of semantic networks to overcome the potential limitations of the quantitatively oriented webometric technique and examine why and how Korean audiences are interconnected.

3.2. Data collection and analysis techniques This study’s sample included messages about the Seoul mayoral TV debates in 2011: three debates organized by three TV networks between October 10 and 13, one debate organized by the Election Broadcasting Debate Commission and broadcasted by three TV networks on October 20, and one debate organized by the Korea Broadcasting Journalists Club and broadcasted by YTN on October 24. More specifically, the SBS TV debate was held on October 10 (Monday) from 20:45 to 21:55, and the KBS TV debate was held on October 11 (Tuesday) from 22:00 to 23:10. After a one-day break, the MBC TV debate was held on October 13 (Thursday) from 23:00 to 00:05. The joint debate organized by the Election Broadcasting Debate Commission was held a week later on October 20 (Thursday) from 23:10 to 00:20 and was jointly broadcasted by KBS, MBC, and SBS. Finally the debate organized by the Korea Broadcasting Journalists Club was held on October 24 (Monday) from 11:00 to 12:00 and was broadcasted by YTN. According to the data on the viewer ratings in the Seoul metropolitan area provided by AGB Neilson Media Research, the ratings and audience shares were 8.3% and 12.5%, respectively, for the SBS debate; 11.0% and 17.0%, respectively, for the KBS debate; and 5.6% and 11.0%, respectively, for the MBC debate. To examine the features of the political discourse on the Seoul mayoral TV debates on SNSs, SNS messages about the debates were collected. Here the following three search terms were employed: ‘‘Seoul mayoral debate,” ‘‘Won-Sun Park debate,” and ‘‘Kyung-Won Na debate” (‘‘서울시장선거”, ‘‘박원순선거”, and ‘‘나경원선거” in Korean). For internet searches based on these terms, those messages that were not related to the study topic were first excluded, and then the top 100 posts were selected using search terms. The most popular 100 posts were selected in terms of the order of search results. To collect data from Twitter and Facebook, the advanced search function in Google was used. The Google search engine was used because of its well-known Page Rank algorithm, which was used to automatically evaluate the hyperlink structure and importance of the Web. In this regard, Google can be regarded as the most representative and reliable data collection channel. For blogs, the blog search function in Naver, the largest portal in Korea, was used. Naver was selected because it was the most popular search engine in Korea. Therefore, blogs written in Korean were well indexed in Naver. Efforts were made to secure a maximum sample size, but the total number of Facebook public posts about the Seoul mayoral by-election was about 300. Lee and Park (2012) examined the Seoul mayoral by-election by focusing on campaign strategies related to Facebook and reported that Facebook made limited public posts available to commercial search engines. In addition, an official bug report on the more recent Korean presidential election documented that performing an API-based search of query terms in Korean is likely to produce a very limited set of results for public posts related to the terms (https://developers. facebook.com/bugs/144346439046252). Therefore, to balance the sample size of the three platforms, the top 100 posts were selected for each search term. The network relationships between terms in messages on Twitter, Facebook, and blogs were analyzed as follows: first, the top 50 most frequently appearing terms were selected for each SNS platform by using the Korean version of KrKwic (Korean Key Words in Context; Park and Leydesdorff, 2004), a semantic network analysis software package developed by Leydesdorff (1995). The semantic network analysis method is a useful content analysis technique that enables researchers to discover symbols, concepts, and meanings from web-mediated documents by extracting words and examining their use patterns (Park and Lee, 2009). Park (2012) examined semantic networks of internet use by Korean students and claimed that this tool can help identify key terms and major issues in text-based responses. A total of 50 terms were selected for each platform because the three platforms imposed different restrictions on the maximum number of words in each post and because 150 terms represented an appropriate size for the analysis. These 150 terms were screened to exclude any overlapping ones and selected a total of 90 terms for the final analysis. Here frequently used terms were assumed to best represent the salience of specific aspects of each candidate from the perspective of the viewertariat. With these 90 terms, sij was developed based on a 3 (three SNS platforms)  90 (words occurring across platforms) matrix S. Here sij, a value between nodes i and j, was used to indicate the frequency of each term for a given SNS platform. In addition, the online social behavior of Korean audiences was measured in terms of their SNS feedback. In this regard, the date and time of each post, the number of comments, hyperlinks, and the frequency of ‘‘likes” (Facebook), hashtags (Twitter), retweets (Twitter), ‘‘sympathy” (blogs; comparable to ‘‘likes” in the Korean blogosphere), and ‘‘view-ons” (blogs; this partly indicates how popular a post is among bloggers) were examined. A social webometric analysis and a social network analysis were conducted using UciNet for Windows (Borgatti et al., 2002), and networks were drawn using the NetDraw function in UciNet.

576

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

4. Results 4.1. Posting frequency and trends by platform Fig. 1 summarizes the posting frequency and trends by platform. Twitter showed the highest posting frequency on October 11 (the date of the KBS debate), followed by October 21 (the day after the Election Broadcasting Debate Commission debate), October 10 (the date of the SBS debate), and October 14 (the day after the MBC debate), in that order. Facebook showed the highest posting frequency on October 11 (the day after the SBS debate), October 14 (the day after the MBC debate), and October 24 (the day after the YTN debate), in that order. Blogs shows the highest posting frequency on October

Fig. 1. Posting frequency and trends by platform.

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

577

14 (the day after the SBS, KBS, and MBC debates), October 21 (the day after the Election Broadcasting Debate Commission debate), and October 11 (the day after the SBS debate), in that order. In terms of changes in posting frequency by platform, Twitter showed a high ratio of real-time posts. As shown in Fig. 1, posting frequency was high during the TV debates (i.e., October 10, 11, 13, 20, and 24). There were 236 Tweets during the TV debates, which accounted for 81.94% of all Tweets. This result indicates that most of the Tweets were sent by viewers who were watching the debates. In terms of Tweets, noteworthy is that a majority of posts concerned the viewers’ prompt opinions on the candidates. Many Tweets were about the attitudes and communication styles of Won-Sun Park and Kyung-Won Na, the two leading candidates. The most frequently retweeted post was ‘‘while I was watching the Kyung-Won Na debate, I felt that she was like a game cock in the arena. She should continue working as a lawyer. She is unqualified as Seoul mayor. . . the Seoul mayoral election is not like choosing an elementary student representative. . .” This Tweet was posted on October 21 at 00:37 and was retweeted 101 times. This Tweet was followed by the post ‘‘only ask closed questions when you debate with KyungWon Na. . . something like. . . did you or did you not go to the Japanese Self-Defense Forces Celebration or did you or did you not use KRW 100 million on cosmetics and spa visits, etc.,” which was posted on October 21 at 00:43 and retweeted 100 times. In terms of Facebook posts, posting frequency was high on days after the debates. As shown in Fig. 1, there were many posts on October 11 and 14 because most posts about the debates were written after midnight (i.e., the next day). Among the 228 posts analyzed, only 33 were posted during the debates, indicating that Facebook users tended to post their comments immediately after watching the debates than during the debates on a real-time basis. This indicates that Facebook was less instantaneous than Twitter in terms of real-time responses. Facebook users tended to comment on the first TV debate, and Tweets during the first TV debate were scattered. As shown in Fig. 1, most posts were made on October 11, which may be due to extra viewer attention given to the first debate (October 10). Unlike Tweets, Facebook posts tended to focus more on discussing concrete and specific issues than on providing a comprehensive evaluation of the debates. For example, when Kyung-Won Na spoke about the bookkeeping of Seoul City based on misinformed facts, there was a sharp increase in the number of posts. The post that received the largest number of likes argued that ‘‘Kyung-Won Na said during the SBS debate that Seoul City uses single-entry bookkeeping, but that is not true. . .” This post was made on October 10 at 18:01 and received 179 likes with 15 ‘‘shares.” In addition, most comments were positive, not argumentative. The post that received the largest number of comments (51, including 7 by the original poster) was written by an assistant in the Kyung-Won Na camp on October 11 at 23:31. These comments focused mainly on the ideology of Won-Sun Park and were generally sympathetic (e.g., ‘‘Won-Sun Park’s policy is like socialism,” ‘‘Won-Sun Park himself is a North Korean follower and a member of the communist group,” and ‘‘conservatives, learn the trickery and agitation skills of the progressives”). Finally, in terms of blogs, many posts were carefully designed. As is shown in Fig. 1, the largest number of posts occurred on October 14, which may be due to the fact that this was the day after the SBS, KBS, and MBC debates. October 21, the day after the Election Broadcasting Debate Commission debate, showed the second-largest number of posts, which may be due the same reason as that for October 14. Among the three types of SNSs, blogs showed the longest time interval between TV debates and posts, and posts occurred at critical moments. Blog posts focused more on providing a comprehensive evaluation of the debates and a professional analysis of campaign pledges than on offering an evaluation of the candidates. The post receiving the largest number of ‘‘sympathy” ratings was ‘‘Won-Sun Park won the Seoul mayoral election. Voters’ choice was Won-Sun Park’s ‘donation,’ not Kyung-Won Na’s ‘skin’!” This post summarized the effects of the debates on the election as well as the general election context and was posted on October 26 at 23:09. This post received 73 sympathy ratings in addition to the largest number of ‘‘view-ons” (121), which represents a major criterion for rating blog exposure. The post with the largest number of comments (2587) was made on October 10 at 22:29 and stated that ‘‘Kyung-Won Na vs. Won-Sun Park TV debate  Kyung-Won Na achieved a complete victory!” The popularity of this post may be due to its provocative title and the heated debate on its extremely pro-Na content. 4.2. An analysis of semantic networks and features A social network analysis was conducted for each platform. Network centrality, which is an index indicating the degree to which a specific node is positioned at the center of the whole network, was first measured. For this, degree centrality, closeness centrality, and betweenness centrality were considered. For an effective analysis, only the top 20 terms (i.e., nodes) were considered. Degree centrality indicates the strength of direct connections (in terms of co-occurrences) between nodes. As shown in Table 2, similar terms appear at the top for each platform. Here the values in parentheses indicate centrality scores of individual terms in the semantic network, and the values in the first column indicate their quartile range. In terms of blog posts, terms such as ‘‘candidate,” ‘‘debate,” ‘‘Won-Sun Park,” ‘‘Seoul,” and ‘‘Kyung-Won Na” showed high degree centrality; for Facebook posts, terms such as ‘‘candidate,” ‘‘Seoul,” ‘‘debate,” ‘‘Kyung-Won Na,” and ‘‘work” showed high degree centrality; and for Tweets, terms such as ‘‘me,” ‘‘debate,” ‘‘Kyung-Won Na,” ‘‘Won-Sun Park,” and ‘‘candidate” showed high degree centrality. Noteworthy is that the term ‘‘me” showed the highest degree centrality for Twitter, indicating that selfassertive first-person expressions such as ‘‘I” sentences were frequently used in Tweets. Closeness centrality measures not only direct connections between nodes but also the indirect distance between nodes. Table 3 shows the results for closeness centrality by platform. In terms of the top 20 blog terms, there were some terms not

578

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

Table 2 Degree centrality by platform. %

Blog

Facebook

Twitter

100

Candidate (47,861), debate (24,287), Won-Sun Park (23,694), Seoul (22,037), Kyung-Won Na (21,940)

Candidate (6593), Seoul (4058), debate (3737), Kyung-Won Na (3051), work (2939)

Me (4477), debate (3821), Kyung-Won Na (2830), Won-Sun Park (2628), candidate (1795)

75

Party (18,031), Seoul City (17,188), mayor (16,704), Seoul mayor (13,055), word (12,154)

Seoul City (2597), mayor (2551), Won-Sun Park (2436), election (2352), word (2198)

Seoul (1774), mayor (1670), Seoul City (1583), word (1540), Seoul mayor (1382)

50

Election (11,486), candidate Won-Sun Park [laughing emoticon] (10,338), first (9546), (9048), candidate Park (7758)

TV debate (1785), Seoul mayor (1734), candidate Park (1596), Grand National Party (1543), vote (1244)

Answer (809), room (613), work (598), candidate Won-Sun Park (483), opponent (438), thinking (437)

25

Grand National Party (7547), [laughing emoticon] (6214), candidate Na (6069), human being (6053), citizen (5976)

Panel discussion (1086), candidate Na (1060), citizen (968), thinking (907), policy (894)

Candidate Kyung-Won Na (407), citizen (364), Seoul mayoral debate (351), today (327)

seen in previous analyses, such as ‘‘mother,” ‘‘aide,” ‘‘national assembly,” and ‘‘battleship Cheonanham,” because these terms were used as opinions, descriptions, and accusations regarding Kyung-Won Na’s slogan ‘‘A Seoul where moms feel happy,” the fact that Na herself was raising a child with a handicap, and photos of people with a handicap. In addition, the frequent terms ‘‘aide” and ‘‘national assembly” were associated with posts about the qualifications of Na by her former aide on his personal blog. The term ‘‘battleship Cheonanham” showed high closeness centrality because Na asked Won-Sun Park whether he believed the statement of the Korean government that North Korea attacked Cheonanham. In the case of Facebook, there were analytical terms related to double-entry bookkeeping (which was mentioned in the first TV debate), such as ‘‘certified public account,” ‘‘policy-related campaign pledges,” ‘‘Democratic Labor Party,” ‘‘financial statements,” and ‘‘principles of accounting.” Tweets included terms that focused mainly on opinions or personal issues such as ‘‘Kyung-Won Na vs. Won-Sun Park,” ‘‘way of talking,” ‘‘special issue,” and ‘‘free meals at school.” On the other hand, there were very few terms related to ‘‘double-entry bookkeeping” on bloggers, and Facebook users’ favorite terms related to ‘‘double-entry bookkeeping” occurred frequently on Twitter. In a similar vein, Twitter users’ favorite terms related to ‘‘opinions or personal issues” were included in the list of Facebook users’ frequently used words. In sum, there were significant differences in degree centrality between blogs and Twitter but not between Facebook and Twitter because blogs had more diversified text messages. Betweenness centrality indicates the level of mediation between nodes such that the higher the level of mediation, the stronger the bridging effect. Table 4 shows the results for betweenness centrality by platform. For all three platforms, the terms that showed high betweenness centrality were found among the top 20 terms. For blogs, terms such as ‘‘KyungWon Na,” ‘‘candidate,” ‘‘Won-Sun Park,” and ‘‘Seoul mayor” were directly related to the Seoul mayoral by-election and showed the same betweenness centrality. In terms of Facebook, the terms ‘‘Seoul,” and ‘‘Seoul City” showed higher betweenness centrality than the names of the candidates, but this difference was not significant. In terms of Twitter, terms such as ‘‘debate,” ‘‘me,” ‘‘Kyung-Won Na,” and ‘‘Won-Sun Park” showed exceptionally high betweenness centrality. Noteworthy is that these terms also showed high degree centrality. A closer examination of betweenness centrality shows that the terms related to elections, including ‘‘Seoul,” ‘‘Seoul City,” ‘‘debate,” ‘‘mayor,” and candidate names, played an important role as a bridge across the semantic network. In other words, meanings were created centered around these terms with high betweenness centrality across platforms. This indicates that these terms occurred most frequently and thus that other terms were connected through these central terms. The density of the network was measured for each platform. This density represents the ratio of the number of existing connections between nodes in a network to that of all potential connections in the network and is simple but useful measure

Table 3 Closeness centrality by platform. %

Blog

Facebook

Twitter

100

Mother (516), aide (515), national assembly (501), battleship Cheonanham (497), suggestion (494)

Certified public account (273), policy-related campaign pledges (271), Democratic Labor Party (270), financial statements (266), street (265)

Kyung-Won Na vs. Won-Sun Park (287), way of talking (287), special issue (287), national bitch (286), free meals at school (286)

75

Announcer (493), Suk-HeeSohn (492), gap (492), resident (490), woman (485)

Letter (264), RohMoo-Hyun (264), local government (264), principles of accounting (264), [laughing emoticon] (263)

Ecosystem (285), hardware (285), Samsung (285), Apple (285), double-entry bookkeeping system (284)

50

Refusal (484), reelection (480), judgment (480), use (480)

Hongik Univ. (255), according (255), watching (255), candidate (255), private residence (254)

[crying emoticon] (283), the weak (283), Won-Sun Park debate (282), honesty (282)

25

Opportunity (475), presidential election (475), era (472), experience (471), female (470), afternoon (470)

Welfare (252), the best way (252), SNU law school (250), alley (250), next year (250)

Capacity as mayor (281), the best way (281), panel (281), snicker (280)

579

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583 Table 4 Betweenness centrality by platform. %

Blog

100

Facebook

Twitter

Seoul (277.30), Seoul City (264.72), debate (222.58), Kyung-Won Na (222.17), work (219.81)

Debate (862.23) me (862.23), Kyung-Won Na (772.88), Won-Sun Park (772.68), me (862.23), mayor (430.53)

75

Kyung-Won Na (79.98), candidate (79.98), Won-Sun Park (79.98), Seoul mayor (79.98), election (79.98), word (79.98), Seoul (79.98), mayor (79.98), Seoul City (79.98), party (79.98)

Mayor (203.24), Won-Sun Park (201.90), candidate (189.76), Seoul mayor (179.38), word (165.06)

Seoul (428.71), Seoul City (428.71), word (423.84), Seoul mayor (372.16), candidate (341.22)

50

First (78.40), Grand National Party (77.83), debate (75.99), candidate Won-Sun Park (75.91), citizen (68.89)

Grand National Party (120.81), vote (112.86), election (108.72), policy (101.16), TV debate (92.76)

Answer (207.45), work (170.98), room (140.05), policy (119.42), mouth (110.47), candidate Won-Sun Park (106.05)

25

TV debate (67.44), policy (66.40), candidate Park (64.68), human being (64.46), problem (62.40)

Thinking (88.16), broadcasting (84.65), October (82.06), yesterday (81.85), candidate Won-Sun Park (78.76)

Citizen (90.66), thinking (78.84), opponent (78.55), candidate Kyung-Won Na (78.46)

of platform cohesiveness. This density was calculated as the number of actual co-occurrences of terms divided by the number of all possible co-occurrences. The latter was calculated by dividing the number (n) of terms by (n 1)/2. In this study, the higher the density, the stronger the cohesiveness of the semantic network was for a given SNS platform. Blogs showed the highest density (14.5), followed by Facebook (4.1) and Twitter (1.8). Fig. 2 shows the results of a comparison of the semantic network of terms between the three platforms. Here the map does not indicate the major terms because its purpose is to clearly visualize the structural differences between the platforms. As shown in the figure, bloggers showed the densest semantic network of terms that were connected to one another. Facebook users also showed dense clusters around central terms. However, Twitterians showed a sparsely connected network, indicating that there were many terms isolated from ones occupying central positions in the map. These results indicate important differences in features between the three platforms. For example, a limited number of issues were extensively discussed on Twitter, which limits the number of words in Tweets, whereas diverse issues were discussed on blogs, which usually feature long posts. Fig. 3 shows a network of 90 major terms from the three platforms. Semantic relationships were visualized after making a co-occurrence matrix of frequently used terms in the sample of posts. Here node size indicates the frequency of terms. Fig. 3 shows the network of major terms from Twitter, Facebook, and blogs. The most frequently occurring terms, namely ‘‘Seoul mayor,” ‘‘Kyung-Won Na,” ‘‘Won-Sun Park,” and ‘‘debate,” were excluded from this network because they were used as search terms. The figure shows the terms ‘‘politics,” ‘‘person,” ‘‘thinking,” ‘‘question,” ‘‘Seoul,” and ‘‘word” at the center of the network for the three platforms. In addition, the result showing the co-occurrence of the term ‘‘Se-Hoon Oh” (former mayor of Seoul) reveals the level of interest in communicating through SNSs. However, Fig. 3 shows slightly different patterns. In terms of the most frequently occurring terms, blogs and Facebook shared the largest number of common terms, whereas blogs and Twitter, the smallest number. The terms most commonly appearing in both blog and Facebook posts referred mainly to political issues that were unrelated to the Seoul mayoral byelection and included the terms ‘‘Democratic Party” and ‘‘Grand National Party,” among others. On the other hand, the terms most commonly appearing in both Facebook and Twitter posts were more directly related to the Seoul mayoral TV debates and included the terms ‘‘SBS,” ‘‘yesterday,” ‘‘comment,” and ‘‘candidate Kyung-Won Na,” among others. In terms of non-overlapping terms by platform, Tweets included terms that focused more on opinions on the TV debates, such as ‘‘smart,” ‘‘feeling,” ‘‘dearth,” and ‘‘logic,” than on the Seoul mayoral by-election itself. In terms of Facebook, the most frequently occurring terms were those associated with Kyung-Won Na’s statements about the bookkeeping method of Seoul City during the first TV debate, including ‘‘single-entry bookkeeping,” ‘‘double-entry bookkeeping,” and ‘‘Seoul National University Law school.” For blogs, the most frequently occurring terms were election-related terms or those associated with the political discourse, including ‘‘opposition party candidate,” ‘‘running for office,” ‘‘campaign pledges,” ‘‘verification,”

Twitter

Facebook

Blog

Fig. 2. A comparison of semantic network structures between three SNSs.

580

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

Fig. 3. A network of major terms on Twitter, Facebook, and blogs.

‘‘reelection,” and ‘‘standpoint,” among others. Tweets were most likely to include terms that were directly related to the TV debates, whereas blog posts were least likely to include such terms. 5. Discussion In this study, the effects of SNSs, a new mode of communication, on the diffusion and restructuring of political discourse on TV were examined by analyzing how SNS users who watched the Seoul mayoral TV debates in October 2011 mediated the debates. The results indicate that the interaction patterns varied according to the structure of the SNS platform. First, as a distinct real-time platform, Twitter showed the largest number of real-time posts during the debates and best reflected the features of the viewertariat suggested in Anstead and O’Loughlin (2009), Anstead and O’Loughlin (2011). Many Tweets were subjective opinions starting with ‘‘I. . .,” and their content focused more on users’ opinions on the debates than on their overall evaluation. The frequently occurring terms in Tweets were related mainly to personal feelings, such as ‘‘nonsense,” ‘‘smart,” ‘‘ridiculous,” and ‘‘slowly,” and many popular retweets were personal in nature. Second, Facebook users were more likely to post messages immediately after the TV debates than during the debates. That is, Facebook users did not post their messages on a real-time basis. In addition, the content of Facebook posts was influenced more by specific and concrete statements from the candidates than by personal impressions. The numbers of likes and shares (which reflect the popularity of posts) received by these posts and the relatively high frequency of terms of such as ‘‘single-entry bookkeeping,” ‘‘double-entry bookkeeping,” ‘‘transfer of political power,” and ‘‘judgment” provide support for this finding. Noteworthy is that Facebook posts were more positive than argumentative. Most studies evaluating the idea that social media reinforce similarities between users have focused mainly on Facebook. Viswanath et al. (2009) examined Facebook discussions on abortion and provided support for this argument, and Choudhury et al. (2010) reported that views of the world and values shared by friends through Facebook can facilitate the sharing and diffusion of similar information. Jiang and de Bruijn (2014) found that Facebook facilitated communication between heterogeneous groups. However, the results of the present study provide a somewhat inconsistent result, indicating that homogeneity was a more important factor than heterogeneity in facilitating political discourse among Facebook users in Korea. Finally, blogs showed the longest time interval between the TV debates and posts and featured many predesigned posts. The number of posts increased sharply the day after the final debate organized by three TV networks and the day after the debate organized by the Election Broadcasting Debate Commission, indicating that the posts were carefully considered and written, not impulse-driven. In terms of content, many posts summarized campaign pledges and debate topics. Like Facebook posts, blog posts showed larger numbers of sympathy ratings and page views for posts summarizing the debates than for those making specific statements about the candidates. In terms of frequency, blog posts included more technical

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

581

terms such as ‘‘verification,” ‘‘opinion survey,” ‘‘reason,” ‘‘judgment,” and ‘‘suspicion” than Tweets or Facebook posts. Unlike Facebook posts, blog posts were more argumentative than positive. For example, the post ‘‘Kyung-Won Na vs. Won-Sun Park TV debate  Kyung-Won Na achieved a complete victory!” received a total of 2587 comments. There were many comments because individual posts were exposed to many people. In addition, this result indicates the presence of various pros and cons and the continuance of a viable debate. According to the semantic network analysis, blogs were much more cohesive than Facebook and Twitter, indicating that blogs facilitated more active discussions on diverse issues than Facebook or Twitter. Although Facebook and Twitter were less cohesive than blogs, the most frequently occurring terms with high degree or betweenness centrality were similar, indicating that discussions on these three platforms focused more on major issues than on peripheral issues. In Korea, where a vertical political culture is a dominant form, major broadcasting stations and newspaper agencies have influenced the formation of public opinion. In the past, there was a clear division between political information suppliers and consumers. Because consumers passively received information only from suppliers, political discourse occurred mainly through the mass media. However, the diffusion of horizontal communication, including the deployment of SNSs, has transformed political and social relationships in Korea in recent years. The surprising victory of Won-Sun Park, an independent newcomer, over the candidate of the ruling conservative party verifies this change in the communication environment. However, this change does not necessarily mean the demise of the mass media’s political influence. Instead, a new type of public sphere for political discourse is being constructed through the interaction between the traditional mass media and various social media platforms. Therefore, SNSs may not push political mobilization toward one specific direction. Instead, political participation and mobilization can vary according to the social media platform. 6. Conclusions This study analyzes and compares the features of the political discourse on the Seoul mayoral TV debates on Twitter, Facebook, and blogs. Few studies have provided an empirical analysis of how users of microblogging sites in Korea mediate political events on TV. More specifically, no study has analyzed and compared SNS platforms such as Twitter, Facebook, and blogs in this context. In this regard, the present study contributes to the literature by investigating the patterns of messages posted by SNS users in Korea while watching the Seoul mayoral TV debates. SNSs are regarded as versatile media that can be used for interpersonal communication as well as for many-to-many communication. In other words, SNSs can transcend the conventional dichotomy between the mass media (disseminating large amounts of information) and telecommunications (facilitating the sharing of private information). More TV viewers are sharing opinions and developing arguments through SNSs on a real-time basis while watching political programs on TV. This means that these viewers not only interpret political discourse while watching TV but also participate in the real-time production and diffusion of the discourse. This suggests that these viewers belong not to a group of embracers but to an active group of individuals who interpret and participate simultaneously (Jenkins, 2006). This indicates a need for new definitions of the social meaning and roles of TV viewers. People before the era of the mass media shared common and instant experiences in one space (Livingstone, 1999). With the advent of the mass media, however, people were no longer bound to the same time and space. Instead, they devolved into a mass of anonymous individuals. However, advances in IT and the diffusion of interactive media have dramatically changed the communication environment. People are no longer anonymous beings, and in particular, the recent global diffusion of social media has enabled the real-time sharing and diffusion of information as well as the real-time sharing of common experiences in one space. In addition, recent years have witnessed the emergence of a new type of pseudo-communal embracing similar to that in the era before the mass media. 7. Limitations and future research This study has some limitations. First, the sample included fewer than 300 posts. Although these posts can be considered credible because the analysis attempted to retrieve all posts related to the TV debates by using a major search engine, any interpretation of the findings should be made with caution. Nonetheless, previous studies of opinion leaders in the blogosphere (Song et al., 2007) and influential Twitter users (Weng et al., 2010) also collected data by using Googlebased methods. Second, some focused interviews were conducted with only nine users to examine their SNS behavior while they watched political programs on TV. But the number of interviewees was too small to justify an empirical analysis of their responses. Third, data were not collected during the TV debates. That is, this study’s data collection method can be viewed as a post hoc evaluation of past events. In this regard, the term ‘‘interaction” in this study may be akin to the term ‘‘two-way communication” in terms of measurement methods. Because this approach may raise some questions, future research should employ a more sophisticated research design to collect big data on a real-time basis to better satisfy the theoretical assumption of this ‘‘interaction” (Vaccari et al., 2015). Finally, this study is descriptive and topological in nature. In this regard, future research should examine whether watching TV with a specific motive can reinforce or change the viewer’s behavior. In addition, the political impact of new communications technologies on public opinion can vary depending on individual characteristics such as sociodemographic factors (Park, 2014) and on the size of networks, including the number of close friends (Jang et al., 2014). Despite these limitations, this study contributes to the literature by being the first to address the question of how SNS platforms such as Twitter, Facebook, and blogs interact with TV in the Korean context. The results are expected to stimulate further research on the relationship between traditional and new media.

582

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

References Ackland, R., 2013. Web Social Science: Concepts, Data and Tools for Social Scientists in the Digital Age. Sage. Ampofo, L., Anstead, N., O’Loughlin, B., 2011. Trust, confidence, credibility: citizen responses on Twitter to opinion polls during the 2010 UK general election. Inform. Commun. Soc. 14 (6), 850–871. Anstead, N., O’Loughlin, B., 2009. Emerging Viewertariat: explaining twitter responses to Nick Griffin’s appearance on BBC question time. UEA School of Political, Social and International Studies Working Paper. Anstead, N., O’Loughlin, B., 2011. The emerging Viewertariat and BBC question time: television debate and real-time commenting online. Int. J. Press/Politics 16 (4), 1–23. Anstead, N., O’Loughlin, B., 2015. Social media analysis and public opinion: the 2010 UK general election. J. Comput. Mediat. Commun. 20 (2), 204–220. http://dx.doi.org/10.1111/jcc4.12102. Barash, V., Golder, S., 2010. Twitter: conversation, entertainment, and information, all in one network! In: Hansen, D., Shneiderman, B., Smith, M.A. (Eds.), Analyzing Social Media Networks with NodeXL: Insights from a Connected World. Morgan Kaufmann, Massachusetts, pp. 143–164. Blood, R., 2005. Weblogs: a history and perspective. In: Bucy, E.P. (Ed.), Living in the Information Age. Wadsworth, Belmont, pp. 129–133. Bond, R.M., Fariss, C.J., Jones, J.J., Kramer, A.D., Marlow, C., Settle, J.E., Fowler, J.H., 2012. A 61-million-person experiment in social influence and political mobilization. Nature 489 (7415), 295–298. Bondad-Brown, B.A., Rice, R.E., Pearce, K.E., 2012. Influences on TV viewing and online user-shared video use: demographics, generations, contextual age, media use, motivations, and audience activity. J. Broadcast. Electron. Media 56 (4), 471–493. Borgatti, S.P., Everett, M.G., Freeman, L.C., 2002. UCINET for Windows: Software for Social Network Analysis. Analytic Technologies, Massachusetts. Bruns, A., 2008. Blogs, Wikipedia, Second Life, and Beyond: From Production to Produsage. Peter Lang, New York. Castells, M., 2009. Communication Power. Oxford University Press, New York. Chadwick, A., 2011a. Britain‘s first live televised party leaders‘ debate: from the news cycle to the political information. Parliamentary Affairs 64 (1), 24–44. Chadwick, A., 2011b. The political information cycle in a hybrid news system: the British Prime Minister and the ‘Bullygate’ affair. Int. J. Press/Politics 16 (1), 3–29. Chang, W.Y., Park, H.W., 2012. The network structure of the Korean blogosphere. J. Comput. Mediat. Commun. 17, 216–230 (a special issue on hyperlinked society). Chin, W.W., Gopal, A., Salisbury, W.D., 1997. Advancing the theory of adaptive structuration: the development of a scale to measure faithfulness of appropriation. Inform. Syst. Res. 8 (4), 342–367. Cho, I.H., Park, J.Y., Park, H.W., 2011. Microblogging political events in South Korea: TV debate between the candidates in the Seoul mayor election. J. Soc. Hum. Stud. East Asia 19, 323–359. Choudhury, M.D., Sundaram, H., John, A., Seligmann, D.D., Kelliher, A., 2010. Birds of a feather: does homophily among users impact information diffusion in social media? (retrieved from ). DeSanctis, G., Poole, M.S., 1994. Capturing the complexity in advanced technology use: adaptive structuration theory. Organ. Sci. 5 (2), 121–146. D’heer, E., Verdegem, P., 2015. What social media data mean for audience studies: a multidimensional investigation of Twitter use during a current affairs TV programme. Inform. Commun. Soc. 18 (2), 221–234. Election Broadcasting Debate Commission, 2010. Election broadcasting debate white paper (retrieved from ). Freelon, D., Karpf, D., 2015. Of big birds and bayonets: hybrid Twitter interactivity in the 2012 presidential debates. Inform. Commun. Soc. 18 (4), 390–406. Giddens, A., 1984. The Constitution of Society: Outline of the Theory of Structuration. University of California Press, Berkeley. Hall, S., 1982. Policing the Crisis: Mugging, the State and Law and Order. Macmillan Press, London. Han, J.W., 2012. Networked Information Technologies, Elections, and Politics: Korea and the United States. Lexington Books. Hansen, D., Shneiderman, B., Smith, M., 2010. Analyzing Social Media Networks with NodeXL. Elsevier, London. Hardt, M., Negri, A., 2004. Multitude: War and Democracy in the Age of Empire. Penguin, New York. Hill, S., Benton, A.D., 2012. Social TV: linking TV content to buzz and sales. Proceedings of 2012 International Conference on Information Systems. Hsu, C.L., Park, H.W., 2011. Sociology of hyperlink networks of web 1.0, web 2.0, and Twitter: a case study of South Korea. Soc. Sci. Comput. Rev. 29 (3), 354– 368. Hwang, Y., Lim, J.S., 2015. The impact of engagement motives for social TV on social presence and sports channel commitment. Telemat. Informat. (available online 2803.15). Jang, S.M., Lee, H., Park, Y.J., 2014. The more friends, the less political talk? Predictors of Facebook discussions among college students. Cyberpsychol. Behav. Soc. Netw. 17 (5), 271–275. Java, A., Song, X., Finin, T., Tseng, B., 2007. Why we Twitter: understanding microblogging usage and communities. Proceedings of the Joint 9th WEBKDD and 1st SNA-KDD Workshop 2007 (retrieved from ). Jenkins, H., 2006. Convergence Culture: Where Old and New Media Collide. New York University Press, New York. Jiang, Y., de Bruijn, O., 2014. Facebook helps: a case study of cross-cultural social networking and social capital. Inform. Commun. Soc. 17 (6), 732–749. Kim, J.H., 2010. The impact of new communication technology on the changing role of traditional news media in the agenda setting process. International Conference on Agenda Setting in Flux, 13–14. Kim, M.K., Kim, K.K., 2004. The use and evaluation of media in voters of university students: based on the results of survey in the 17th general election. Korean J. Broadcast. 18 (3), 134–174. Korea Internet & Security Agency, 2012. Survey on the Internet Usage (retrieved from ). Larsson, A.O., 2012a. Tweeting the viewer: use of twitter in a talk show context. A Paper Presented at the 2012 AoIR (Association of Internet Researchers) Conference. Larsson, A.O., 2012b. Understanding non-use of interactivity in online newspapers: insights from structuration theory. Inform. Soc. 28 (4), 253–263. Leavitt, A., Burchard, E., Fisher, D., Gilbert, S., 2009. The influential: new approaches for analyzing influence on Twitter. Web Ecology Project (retrieved from ). Lee, Y.O., Park, H.W., 2012. What do Facebookers say? The 2011 Seoul mayoral by-election in social media. A Paper Accepted for the Panel of ‘Political Participation in the Web 2.0’, World Congress of Political Science, Madrid, Spain (retrieved from ). Leydesdorff, L., 1995. The Challenge of Scientometrics: The Development, Measurement, and Self-organization of Scientific Communications. DSWO Press, Leiden (retrieved from ). Lim, Y.S., Park, H.W., 2011. How do congressional members appear on the web? Tracking the web visibility of South Korean politicians. Gov. Inform. Q. 28 (4), 514–521. Lim, Y.S., Park, H.W., 2013. The structural relationship between politicians’ web visibility and political finance networks: a case study of South Korea’s National Assembly members. New Media Soc. 15 (1), 93–108. Lin, C.A., 2011. Media substitution: supplementation, complementarity, or displacement? In: Vishwanath, A., Barnett, G.A. (Eds.), The Diffusion of Innovations: A Communication Science Perspective. Peter Lang, New York, pp. 171–192. Livingstone, S., 1999. New media, new audience? New Media Soc. 1 (1), 59–66. Lull, J., 1990. Inside Family Viewing: Ethnographic Research on Television’s Audiences. Routledge, London. McCombs, M.E., Shaw, D.L., 1972. The agenda-setting function of mass media. Public Opin. Q. 36 (2), 176–187. Newcomb, H., 1994. Television: The Critical View. Oxford University Press, New York, NY. Page, R., 2012. Re-examining narrativity: small stories in status updates. Text Talk 30 (4), 423–444.

Y.-C. Heo et al. / Telematics and Informatics 33 (2016) 570–583

583

Park, H.W., 2010. Mapping the e-science landscape in South Korea using the webometrics method. J. Comput. Mediat. Commun. 15, 211–229. Park, H.W., 2012. Examining academic Internet use using a combined method. Qual. Quantity 46 (1), 251–266. Park, Y.J., 2014. My whole world’s in my palm! The second-level divide of teenagers’ mobile use and skill. New Media Soc. http://dx.doi.org/10.1177/ 1461444813520302. Park, H.W., Jankowski, N., 2008. A hyperlink network analysis of citizen blogs in South Korean politics. Javnost Public 15 (2), 57–74 (a special issue on methodological issues in conducting online political communication research). Park, H.W., Lee, Y.-O., 2009. A mixed text analysis of user comments in a portal site: ‘BBKgate’ in the 2007 presidential election of South Korea. J. Korean Data Anal. Soc. 11 (2), 731–744 (written in Korean). Park, H.W., Leydesdorff, L., 2004. Understanding the KrKwic: a computer program for the analysis of Korean text. J. Korean Data Anal. Soc. 6 (5), 1377–1388. Park, S.J., Lim, Y.S., Sams, S., Sang, M.N., Park, H.W., 2011. Networked politics on cyworld: the text and sentiment of Korean political profiles. Soc. Sci. Comput. Rev. 29 (3), 288–299. Pressgrove, G., 2012. Top U.S. nonprofits use of web 2.0 to achieve dialogical potential. A Paper Presented at the Social Media and PR Session, International Communication Association, Phoenix, Arizona, U.S. Proulx, M., Shepatin, S., 2012. Social TV: How Marketers can Reach and Engage Viewers by Connecting Television to the Web, Social Media, and Mobile. Wiley & Sons, Hoboken, NJ. Rubin, A.M., 1981. An examination of television viewing motivations. Commun. Res. 8, 141–165. Sams, S., Lim, Y.S., Park, H.W., 2011. E-research applications for tracking online socio-political capital in the Asia-Pacific region. Asian J. Commun. 2 (5), 450– 466. Shih, C., 2009. The Facebook Era: Tapping Online Social Networks to Build Better Products, Reach New Audience, and Sell More Stuff. Prentice Hall, Boston. Shirky, C., 2008. Here Comes Everybody: The Power of Organizing Without Organizations. Penguin Press, New York. Shirky, C., 2011. The political power of social media. Foreign Affairs 90 (1), 28–41. Skoric, M.M., Park, Y.J., 2014. Culture, technologies and democracy: a cross-national analysis of political development. Telemat. Informat. 31 (3), 364–375. Smith, A., Boyles, J.L., 2012. The Rise of the ‘Connected Viewer’. Pew Internet & American Life Project (retrieved from ). Song, X., Chi, Y., Hino, K., Tseng, B., 2007. Identifying opinion leaders in the blogosphere. Proceedings of the sixteenth ACM conference on Conference on information and knowledge management, pp. 971–974 (ACM). Stefanone, M.A., Lackaff, D., Rosen, D., 2010. The relationship between traditional mass media and ‘‘social media”: reality television as a model for social network site behavior. J. Broadcast. Electron. Media 54 (3), 508–525. Thelwall, M., 2009. Introduction to Webometrics. Morgan & Claypool, New York. Vaccari, C., Valeriani, A., Barberá, P., Bonneau, R., Jost, J.T., Nagler, J., Tucker, J.A., 2015. Political expression and action on social media: exploring the relationship between lower- and higher-threshold political activities among Twitter users in Italy. J. Comput. Mediat. Commun. 20 (2), 221–239. http:// dx.doi.org/10.1111/jcc4.12108. Viswanath, B., Mislove, A., Cha, M., Gummadi, K.P., 2009. On the evolution of user interaction in Facebook. Proceedings of the 2nd ACM SIGCOMM Workshop on Social Networks WOSN. Webster, J., 2011. The duality of media: a structurational theory of public attention. Commun. Theory 21 (1), 43–66. Weimann, G., Tustin, D.H., van Vuuren, D., Fouber, J.P.R., 2007. Looking for opinion leaders: traditional vs. modern measures in traditional societies. Int. J. Public Opin. Res. 19, 173–190. Weng, J., Lim, E.P., Jiang, J., He, Q., 2010. Twitterrank: finding topic-sensitive influential twitterers. Proceedings of the Third ACM International Conference on Web Search and Data Mining, pp. 261–270 (ACM). Wohn, D.Y., Na, E.K., 2011. Tweeting about TV: sharing television viewing experiences via social media message streams. First Monday 16 (3), 1–29. Zhao, D., Rosson, M.B., 2009. How and why people Twitter: the role that micro-blogging plays in informal communication at work. Proceedings of the ACM 2009 International Conference on Supporting Group Work, pp. 261–270 (retrieved from ).

News Articles UK election online fundraising (CNN, 4 May, 2010) (retrieved from ). Web 2.0: the new election superweapon (The Guardian, 11 April, 2010) (retrieved from ). South Korea smart phone subscribers exceed 20 million (Reuters, 30 October, 2011) (retrieved from ).