Detecting collaboration patterns among iSchools by linking scholarly communication to social networking at the macro and micro level
1 like504 views
The document investigates collaboration patterns among iSchools by linking aspects of scholarly communication and social networking at both macro and micro levels. It poses three research questions concerning the interests, research topics, and relationship between social networking and scholarly communication of iSchools. Methodologies include data collection from Twitter and conference proceedings, employing various network analysis techniques such as QAP correlation and pathfinder algorithms.
Detecting collaboration patterns among iSchools by linking scholarly communication to social networking at the macro and micro level
- 1. Detecting collaboration patterns among iSchools by linking scholarly communication to social networking at the macro and micro level So Young YU Dept. of Library and Information Science Hannam University soyoungyu201@gmail.com
- 2. 2
- 3. Research Questions • RQ1. What are the major interests and concerns of iSchools in social networking? • RQ2. What are the major research topics of iSchools in scholarly communication? • RQ3. Is there a relationship between social networking and scholarly communication of iSchools at the macro and micro levels? 3
- 4. Methodology 4 Data collection Pre-processing and network extraction Micro-level analysis Macro-level analysis
- 5. Data Collection • 41 iSchools from iSchool Directory of iSchool Organization http://ischools.org/directory/ [accessed April 10th 2013) • 16494 accounts of Ischools’ twitter and their followees – by NodeXL v.1.28 – June 17th, 2013 • 749 Papers and posters published in iConference proceedings – 2008-2012 – IDEALS https://www.ideals.illinois.edu/handle/2142/14872 – ACM Digital Library http://dl.acm.org/event.cfm?id=RE615&CFID=233110227& CFTOKEN=50918548 5
- 6. Pre-Processing and network extraction Twitter data Identifying unique iSchools and their followees accounts Indexing description Following network (16,494 X 16,494) iConference data Unifying names of iSchools and non-iSchools Indexing author keywords Document - Indexed author keyword network (749 X 1,520) Document - Institution Network (749 X 196) 6
- 7. Institution Profiling Matrix (iSchools X Documents ) X (Keywords X Documents) 𝑇 = (iSchools X Keywords) • ‘Author Profile Vector’ (Kim & Lee 2007) 7
- 8. Level of analysis with pertinent networks 8 Micro-level (40 iSchools) • N1. Co-following network • N2. Co-authorship network • N3. Institution profiling network Macro-level (iSchools and ‘Non’- iSchools) • N4. Co-word network of top 1% followees’ description • N5. Co-word network of author keyword • N6. Co-authorship network • N7. Institution profiling network
- 9. RQs with pertinent networks RQ1 • N1 • N4 & N5 RQ2 • N2 & N3 • N6 & N7 RQ3 • QAP (N1,N2,&N3) • Comparison (N4& N5) • QAP (N6&N7) 9
- 10. Methods in Network analysis • PathFinder Algorithm for visualization • QAP (Quadratic Assignment Procedure) Correlation for comparison 10
- 11. Level of analysis with pertinent RQs and networks RQ1 RQ2 RQ3 Micro-level N1 N2 & N3 QAP correlation among N1,2, & 3 Macro-level N4 & N5 N6 & N7 Comparison between N4 & N5 QAP correlation between N6& N7 11
- 12. RESULT 12
- 13. Pathfinder Network of Co-following at micro-level 13
- 14. Collaboration Network at micro-level 14
- 15. Pathfinder Network of institution profiling at micro-level 15
- 16. Comparison between Social Networking and Scholarly Communication N1 N2 N3 N1 1.00 N2 0.06** 1.00 N3 0.41** 0.24 1.00 16 QAP-correlation at micro-level analysis Significant correlation between iSchools’ interest in current issues and their research topics
- 17. Pathfinder Network of top interest in social networking 17
- 18. Pathfinder Network of top interest in research 18
- 19. collaboration and topical relevance :macro level Scholarly communication Patterns at macro-level ((a) Coauthorship (b) Institution Profiling) 19 QAP-correlation (r = .39, p <.01)
- 20. collaboration and topical relevance :micro-level 20 QAP-correlation (r = .24, p >.01)
- 21. Conclusion • ‘Candidates of Brokers’ in collaboration in research • ‘Candidates of Brokers’ in sharing research topics in common • Active in social networking and scholarly communication, whilst inactive in research collaboration • Follow-ups 21
- 22. 22
Editor's Notes
- #15: size of node = betweeness centrality , the number in parenthesis = degree centrality