The Internet, Policy & Politics Conferences

Oxford Internet Institute, University of Oxford

Yoonjae Nam, Yeon-Ok Lee, Han Woo Park: An analysis of web ecology examined through social webometrics during 2012 general election in south Korea

Yoonjae Nam, Kyunghee University
Yeon-Ok Lee, Royal Holloway, University of London
Han Woo Park, Yeungnam University


This study tries to examine the inter-relationship among websites by the human activity of using the Internet in information ecology. For the ecological perspective, integrations and changes of diverse information behavior by people over time should be observed and described. Based on literature reviews, this study attempts to provide the description of flow and use of information—“web ecology”—during the campaign period of the general elections on April 11, 2012 in South Korea. Information influenced by a dynamic situation includes voters’ interests in elections, the trend of public opinion, and the appearance of controversial campaign issues, and these affect web ecology. It is assumed that e-campaign practices in South Korea have always provided useful data for global comparison as South Koreans were early adopters of e-campaigning and furthermore they have recently responded in innovative ways to the authorities’ restrictive interpretations and application of pre-Internet election laws to online activity. One phenomenon is that an increasing number of voters have now taken their electoral engagement to international Web 2.0 platforms to avoid domestic legal complication (Lee & Park, 2010). This encourages us to examine the election in order to understand whether, how and to what extent the ‘web ecology’ reflects the ongoing development in the field, or even predicts how an election unfolds.

Research Questions

In this study, we will examine the inter-relationship among websites, measured by the human activity of using the Internet in web ecology during the campaign period of the 2012 regional elections in South Korea. Web ecology does not simply cling to specific action or event by an individual candidate, political party, or voter. It is assumed that examining web ecology helps us to get a sense of relationship between political events and people’s informational behavior with the Internet, since it provides a broad description of integrations and changes in diverse information behavior by people over time. In other words, information in web ecology is influenced by dynamic situations, such as voters’ interests in elections, the trend of public opinion, and the appearance of controversial campaign issues.

RQ1: Can people’s information behaviors during election campaigns be described effectively by examining web ecology?

Web ecology can be measured in various ways, including link structure, sentiment analysis, readership data, conversational structure, topic classification, and temporal analysis (Finn et. al, 2007). Among the techniques to measure web ecology, we examine both “hyperlinking co-occurrences” and “name mentions.”

In addition, web ecology is created by human behavior. There should be people to preserve the ecology itself—a keystone species—who can serve as information producers, translators, facilitators, or mediators. In this study, we attempt to investigate which type of peoples’ informational behavior for political election campaigns is an essential contribution for maintaining web ecology. For this, we compare two types of informational behavior with the Internet: 1) producing and disseminating information (using news articles, blog posts, and twitter messages) and 2) referring information via hyperlinks.

RQ2: How do people serve to preserve a web ecology during political election campaigns?

Data Collection and Analysis

The web ecology between two candidates is operationally defined as the number of both ‘hyperlinking cooccurrences’ and ‘name mentions’. These data can be collected at multiple points during the election period to identify dynamically changing properties of election web ecology.

First, Web-ecology can be examined through network analysis. For this study, hyperlinking coocurrence will be measured with colink data, which refer to incoming links shared between any pair of candidates’ websites (Park, 2010). More specifically, we will retrieve external webpages sending at least one link to a pair of candidate sites under investigation, using a Yahoo’s colink search option. In other words, two candidates are regarded as ‘connected’ if their homepages are linked by Internet users. Note that the idea of using hyperlink concurrence data to quantify the relative level of ‘perceived’ public awareness of political actors and social events was frequently presented before in social webometrics (Thelwall, 2009). The hyperlink research methods have been expanded to understand current political phenomena in special, for example, hyperlink analysis of websites and SNS sites of South Korean National Assembly members (Hsu & Park, 2011), co-link analysis of candidates and their parties during the 2007 Korean presidential election campaign (Park, 2012), and the trackback link network analysis of news blogger during the 2008 candlelight protest in Korea (Chang & Park, 2012).

Based on the results obtained from Yahoo, we construct a co-link network diagrams of the frequently retrieved candidates’ sites during campaigns. In the visual mapping, the number of the lines between sites represent network richness. Also, the size of the lines between sites is proportionally thicker according to the number of external websites co-linking to the sites. The visual mapping has following procedure; 1) node sizes are drawn bigger in accordance with their indegree centralities, 2) line sizes are made thicker based on the frequency of shared incoming links, and 3) isolates are excluded. In addition, we calculate network density by the number of co-links is divided by the number of candidates.

Second, to observe peoples informational behavior during campaigns, name mention is measured with the number of webpages retrieved from the Korea’s top search engine when each candidate’s name is entered as the search query. The web mentions of individual candidates’ names are collected from blogs and news articles. While blog tend to denote online personal media with reflections and comments, news represents corporate channel (Park & Thelwall, 2008). Additionally, name mention data are harvested from twitter sphere using a Korean twitter search service Therefore, measuring a worm of mouth activity on three different spheres is complimentary and particularly useful for the study of social and political structure on the web.


Lee Y-O, Park HW (2010) The Reconfiguration of E-Campaign Practices in Korea: A Case Study of The Presidential Primaries of 2007. International Sociology 25(1): 29-53.

Finin T, Joshi A, Kolari P, Java A, Kale A, Karandikar A (2007) The information ecology of social media and online communities. AI Magazine 28(3).

Park HW (2010) Mapping the e-science landscape in South Korea using the webometrics method. Journal of Computer-Mediated Communication 15(2), 211-229.

Park HW (2012) How do social scientists use link data from search engines to understand Internet-based political and electoral communication. Quality & Quantity 46(2): 679-693.

Thelwall M (2009) Introduction to Webometrics. New York: Morgan & Claypool.

Park HW, Thelwall M (2008) Developing network indicators for ideological landscapes from the political blogosphere in South Korea. Journal of Computer-Mediated Communication 13(4): 856-879.

Hsu C-L & Park HW (2011) Sociology of Hyperlink Networks of Web 1.0, Web 2.0, and Twitter: A Case Study of South Korea. Social Science Computer Review 29 (3): 354-368.