Download - Studying Social Science Using E Tools
STUDYING SOCIAL SCIENCE USING E-TOOLS: CASES OF SOUTH KOREA & TAIWAN
Chien-leng Hsu
WCU Webometrics InstituteYeungNam University
ABOUT THIS PRESENTATION
E-research Development in Social Science Research examples WCU Webometrics Institute
Case 1: Network of politicians in Taiwan Case 2: Network of MPs in South Korea
WHAT IS E-RESEARCH?
A minor but growing approach
The use of new digital tools
E-research in Social Science development of online
tools to automate the research process
Experimentation with new types of data visualisation
Automation of the research process Communicion Research mnagement Data collection & analysis Publication software
Data visualisation Social network analysis Hyperlink analysis Multimedia & dynamic representation
DEVELOPMENT IN SOCIAL SCIENCE (1/2)
Gary King (Harvard University, Miller Converse Lecture Series talk, 2009)
The next 50 years: spectacular increase in new data source Improved, expanded & applied
In depth studies of individual places, people, events Aggregate government statistics Survey research
Data everywhere – the growing Internet The replication movement – academic data
sharing Advances in statistical methods, informatics &
software
DEVELOPMENT IN SOCIAL SCIENCE (2/2)
Examples Opinions of activists
Sample of a few thousand interviews ↪ millions of political opinions every day in
blogosphere Social contacts
Asking respondents to recall names over past years↪ emails, SMSs, social media connections, etc
Progress in the new data-rich world Large-scale, interdisciplinary research Computer-assisted & quantitative New statistical methods & engineering required
NEW TYPES OF DATA (EXAMPLES)
Unstructured text: emails, web pages Geographic location: mobiles Social media: facebook, twitter, virtual
worlds Web surfing artifacts: clicks, searches Government data: electronic data Scholarly data
RESEARCH AREAS (EXAMPLES)
Semantic context Modern text mining approaches to
ancient texts E-linguistics Data management/digital archive Visualisation of dialectology Digital arts & humanities
WCU WEBOMETRICS INSTITUTE
ProjectInvestigating Internet-based Politics with E-Research Tools
Objects To identify, track & analyse the effectiveness of networked political
campaigns across a range of web platforms in Korea To develop software tool
Major areas Semantic Networks on Political Webosphere Cyworld Twitter Korean Internet Network Miner (tool)
Website
http://english-webometrics.yu.ac.kr/
NETWORK OF POLITICIANS IN TAIWAN
Case 1
BRIEF BACKGROUND INFORMATION
Progressive groups use more alternative media
Political struggle between two political camps Conservative: pro-unification with China (Leader:
KMT) Progressive: pro-independence (Leader: DPP)
Approx 80% of mainstream media is biasd toward the conservatives
Research targets: Politicians Plurk
Microblogging sites:a mini form of traditional blogs. Microblogging allows users to send brief messages
WHAT IS PLURK
A popular microblogging site in Taiwan No. of Taiwanese user - 249,441 (July 2009)
03/09
5K
10K 10K
25K
RESULTS & FINDINGS
•Progressive politicians use more Plurk• Progressives: 37 politicians• Conservative: 7 politicians
Age Karma profile viewsfriends invited
no. of plurksno. of plurk
responsesno. of friends no. of fans
ALL(N=44)
46.80 64.77 10719.70 12.34 261.52 2158.30 1783.75 580.05
Progressive(Np=37)
46.11 67.25 12017.59 13.59 291.78 2417.35 2065.68 650.46
Conservative(Nc=7)
50.43 51.71 3859.43 5.71 101.57 789.00 293.57 207.86
Question:
DOES AGE MATTER?
AGE:PROGRESSIVES VS CONSERVATIVES
Age Freq Freq% CumFreq CumFreq%
21-30 1 2.70 1 2.70
31-40 12 32.43 13 35.14
41-50 12 32.43 25 67.57
51-60 7 18.92 32 86.49
61-70 4 10.81 36 97.30
71-80 1 2.70 37 100.0
Sum 37 100.00
Age Freq Freq% CumFreq CumFreq%
21-30 0 0 0 0
31-40 0 0 0 0
41-50 4 57.14 4 57.14
51-60 3 42.86 7 100.00
61-70 0 0 7 100.00
71-80 0 0 7 100.0
Sum 7 100.00
Pearson Correlation
Gender Party Age Plurks
Gender 1.000 -0.329 0.204 0.005
Party -0.329 1.000 -0.156 0.214
Age 0.204 -0.156 1.000 -0.258
Plurks 0.005 0.214 -0.258 1.000
Progressive Politicians Conservatives Politicians
EXCEPTION: HSIEH & SU
Age Karma profile views friends invited no. of plurks no. of plurk responses no. of friends no. of fans
ALL(N=44)
46.80 64.77 10719.70 12.34 261.52 2158.30 1783.75 580.05
Hsieh 63 97.32 65462 100 506 7267 10869 3393
Su 62 96.41 60444 90 405 631 9853 3708
Hsieh
Su
Profile views
Number of friends
Number of fans
DISCUSSIONS
Hsieh & Su: Candidates for the presidential election 2008
Progressive politicians use Plurk Express their opinion Cultivate new constituencies Link to opinion leaders of Taiwanese grassroot
groups The ‘rich get richer’ phenomenon
Plurkers that are the target of many visitors are disproportionately likely to be targeted by any new visitors
Case 2
NETWORK OF MPS IN SOUTH KOREA
WEB 1.0, WEB 2.0 &TWITTER
Research purpose:To investigate structural changes in hyperlink networks from Web 1.0 to Web 2.0 in South Korean Politics
Units of analysis: Congress members of South KoreaYear of observations:
Web 1.0: homepage, 2000 & 2001 Web 2.0: blogs, 2005 & 2006 Twitter: 2009
Blue: GNP: Conservative: Opposition
Red: MDP: Liberal: Ruling
WEB 1.0: HOMEPAGE 2000 VS 2001
Star networks without any isolation
WEB 2.0: BLOGS 2005 VS 2006
2005 2006
WEB 1.0, WEB 2.0 &TWITTER
Web Types YearSum of links
(Mean)
Dens-ity
CentralisationGini
Coeffi-cientIN OUT
Web 1.0(Homepage
)
2000N=24
5
373(1.52) 0.006 1.84 69.33 0.984
2001 515(2.10) 0.009 1.19 99.55 0.996
Web 2.0(Blog)
2005N=99
652(6.59) 0.067 22.07 41.66 0.759
2006 589(5.95) 0.061 20.67 35.10 0.763
Twitter 2009 111(5.05) 0.240 24.72 39.68 0.408
WEB 1.0, WEB 2.0 & TWITTER
Web 1.0: Hub, but sparse network Web 2.0: Hub disappearing, but
becoming dense Twitter: similar to Web 2.0 structure,
and denser More to work (example):
To compare top 10 politicians ego-networks and investigate how they change
CONCLUSIONS
Previous research: Results from web data are informative The Web reflects situations in the physical
world The validity of the results need to be
supported by various qualitative studies Further research:
Previous study areas are still ongoing Cyber emotions: online image content
analysis of web 2.0 politics
THANK YOU
AcknowledgmentsWCU Webometrics Institute acknowledges that this research is supported from the WCU project - investigating internet-based
politics using e-research tools granted from South Korean Government