366 jevin west

311 views

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
311
On SlideShare
0
From Embeds
0
Number of Embeds
14
Actions
Shares
0
Downloads
1
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

366 jevin west

  1. 1. The Eigenfactor Metrics: a networkapproach to assessing scholarly journals Jevin D. West University of Washington jevinw@u.washington.edu
  2. 2. The Eigenfactor Algorithm Q: How can we better evaluate the scholarly literature?The Eigenfactor Project Q: How does network structure affect function?
  3. 3. Citation Networks, Scaling Up Network Effects, Centrality Eigenfactor, Article Influence Mapping, Other Metrics
  4. 4. Citation NetworksCitation Networks, Scaling Up
  5. 5. Citations
  6. 6. Citations
  7. 7. Scholarship is all about the flow of ideas
  8. 8. Networks are all aboutthe flow of information
  9. 9. Citations form a vast networkde Solla Price, Science (1965)
  10. 10. Citations among AuthorsEugene Garfield Derek de Solla Price
  11. 11. Citations among JournalsPhysical Review Letters Ecology Letters
  12. 12. What happens when you scale up? 12,000 Journals 150,000,000 Citations
  13. 13. How do we identify the important journals? …using only the citation network  …but all of the citation network 
  14. 14. Network EffectsNetwork Effects, Centrality
  15. 15. High School Dating Network Image Courtesy of Mark Newman
  16. 16. High School Dating NetworkSystem Scale: Network Topology Matters Q: Which network would MONO prefer? Individual Scale: Connection Source Matters Q: Which girl/boy should I date if I don’t want MONO?
  17. 17. Who is the most ‘influential’ high school student? Image Courtesy of Mark Newman
  18. 18. Degree Centrality Eigenvector Centrality
  19. 19. Degree CentralityLinks 9  5  4  3  2  1  1        2        3         4         5          6        7          8        9     10     11      12     13      14       Number of students
  20. 20. Accounting for the Network
  21. 21. The Network Matters
  22. 22. This network property of the scholarly literature was largely ignored over the first century of scholarly evaluation.How can we extract this information in order to better measure information flow?
  23. 23. Impact factor Garfield, Science (1955)Cites in 2010 to articles in 2009 or 2008 Number of articles in 2009 and 2008
  24. 24. Impact Factor2010 3 citations 2008 - 2009 3 IF2010 = 2 2 articles
  25. 25. Impact Factor Eigenfactor
  26. 26. Citation Networks, Scaling Up Network Effects, Centrality Eigenfactor, Article Influence Mapping, Other Metrics
  27. 27. Eigenfactor algorithmEigenfactor – taking into account the network
  28. 28. 25 nodes and 42 unweighted, undirected links
  29. 29. 25 nodes and 42 weighted, undirected links
  30. 30. Which node is the most central? 22  4  23  2  11  21  9  18  20  7  6  12  3  1  13  10  5  24  14  8  19  17  25  15  16 25 nodes and 42 weighted, directed links
  31. 31. Deterministic: iterative votingStochastic: random walker Applet: Daniel Edler
  32. 32. Eigenfactor – Top Ten Journals in Science (2008) Rank Journal Eigenfactor 1 Nature 1.76 2 PNAS 1.70 3 Science 1.58 4 Journal of Biological Chemistry 1.33 5 Physical Review Letters 1.28 6 Journal of the American Chemical Society 0.95 7 Physical Review B 0.77 8 Applied Physics Letters 0.72 9 New England Journal of Medicine 0.68 10 Cell 0.67 Total number of journals ranked = 8,232
  33. 33. Eigenfactor – Top Ten Journals in Science (2008) Rank Journal Eigenfactor 1 Nature 1.76 % 2 PNAS 1.70 % 3 Science 1.58 % 4 Journal of Biological Chemistry 1.33 % 5 Physical Review Letters 1.28 % 6 Journal of the American Chemical Society 0.95 % 7 Physical Review B 0.77 % 8 Applied Physics Letters 0.72 % 9 New England Journal of Medicine 0.68 % 10 Cell 0.67 % Total number of journals ranked = 8,232
  34. 34. Eigenfactor Article Influence = # articles Eigenfactor, Ranking Mapping
  35. 35. Article Influence – Top Ten Journals in Science (2008) Rank Journal Article Influence 1 Review of Modern Physics 24.9 2 Annual Review of Immunology 24.7 3 Annual Review of Biochemistry 20.9 4 Nature Reviews Molecular Cell Biology 20.0 5 Annual Review of Neuroscience 18.9 6 Cell 18.9 7 New England Journal of Medicine 18.8 8 CA – A Cancer Journal for Clinicians 17.5 9 Nature 17.3 10 Science 16.3 Total number of journals ranked = 8,232
  36. 36. Article Influence vs. Impact Factor Impact Factor, Eigenfactor
  37. 37. Ecology and Evolution (2007)
  38. 38. Impact Factor 0.5  0.1  0.1  0.1  0.1  0.1 
  39. 39. Eigenfactor 0.342  0.316  0.147  0.103  0.056  0.036 
  40. 40. Mapping Mapping, Other Metrics
  41. 41. Slavic Studies Art Education Philosophy Political Science Asian Studies Literature Education Law Language Teaching Management Art and Architecture Marketing Sociology History Mathematics Economics Classical Studies Demography StatisticsBiblical Studies Geography Biology Anthropology Folklore Plant Cell
  42. 42. 1995 2004
  43. 43. Alluvial diagrams Rosvall  Bergstrom (2010) 
  44. 44. Cost-effectiveness
  45. 45. Citations among AuthorsEugene Garfield Derek de Solla Price
  46. 46. Author-level Metrics
  47. 47. Citations vs Downloads
  48. 48. Scholarly Feedback
  49. 49. Three things to remember…1. The scholarly literature forms a vast network 2. There is a wealth of information in the structure of networks 3. Eigenfactor is a network metric that ranks scholarly journals
  50. 50. Acknowledgements Carl Bergstrom Daril Vilhena Martin Rosvall Ben Althouse Ted Bergstrom Johan Bollen Ralph Dandrea Gail Dykstra Daniel Edler Alan Yen Gregg Gordon Michael Jensen Michael Kurtz Moritz Stefaner Theresa Portzer NaEonal Science FoundaEons  NaEonal InsEtutes of Health 

×