Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Approachable Network Analysis

Unlock the power of network structures in your data. Learn how to build and analyze networks to gain insights through relationship analysis. Apply approachable techniques and free, user-friendly software. Transform the data you have into the data you need – from relational databases and unstructured text to common network structures.

Jeff detailed his work in the Medical School Grant Review & Analysis Office. Examples will include: Identifying networks of collaborators from eResearch Proposal Management [eRPM PAF] data, discovering networks of concepts in unstructured text, and use cases from other administrative data sets. Jeff’s presentation included:

-“Networks 101″ – The basic building blocks of networks
-How people in any business unit can apply network analysis
-An emphasis on approachable techniques and free, user-friendly software
-Strategies for effectively visualizing and sharing network-driven insights
-Tools, tips, and tricks

  • Login to see the comments

  • Be the first to like this

Approachable Network Analysis

  1. 1. Approachable Network Analysis Jeff Horon
  2. 2. Gartner’s Hype Cycle Source:
  3. 3. My Mission – Short Circuit the Hype Cycle Source:
  4. 4. You will leave here with the knowledge, skills skills, resources, motivation, and ideas you need to do t d network analysis k l i today with data you probably already have
  5. 5. [Social] Network AnalysisSo,So like Facebook? Sort of of.But tB t networks are everywhere. k hAnd they aren’t necessarily “social.”
  6. 6. TopicsNetworks 101Your Use CasesTransforming YT f i Your D tDataFree, User-Friendly SoftwareExamplesQ&A
  7. 7. Networks 101Building BlocksPutting the Pieces Together – VisualizationMetricsM ti
  8. 8. Building BlocksNodes [Vertices] – People Things Ideas People, Things,Links [Edges] – Relationships or
  9. 9. Visualization
  10. 10. Metrics – Degree Highest Degree
  11. 11. Metrics – Degree – In-Degree Highest g In-Degree “Popular” Popular
  12. 12. Metrics – Degree – Out-Degree Highest Out-Degree g “Gregarious”
  13. 13. Metrics – Betweenness Highest Betweenness “Bridge” “Commonalities”
  14. 14. Metrics – Betweenness
  15. 15. Metrics – Closeness Highest Closeness “Who could spread a rumor?”
  16. 16. Metrics – Closeness
  17. 17. Metrics – Eigenvector Centrality Highest Eigenvector Centrality “Importance”
  18. 18. Metrics – Eigenvector Centrality
  19. 19. Recap Degree ( di D (undirected): N b of d) Number f connections In- / Out-Degree (directed): “Popular” / “Gregarious” “G i ” Betweenness: “Bridges” / “Commonalities” Closeness: “Rumor starting point” Eigenvector Centrality: “Importance”
  20. 20. Your Use Cases – Connect:People to Other PeopleThings/Ideas to Other Things/IdeasPeople to Things/Ideas
  21. 21. If the other attendees are starting to looklike this to you…
  22. 22. Transforming Your DataCommon Network Data StructuresRelational DatabaseUnstructured TextU t t dT t
  23. 23. Edge ListA list of edges (links)!A BA CB C
  24. 24. Edge ListA list of edges (links)!A B A BA C A CB C B C
  25. 25. Edge ListA list of edges (links)!A B A BA C A CB C B C A B C
  26. 26. Data You May Already HaveFaculty/Staff and Appointing DepartmentsFaculty/Staff and GroupsPrincipal Investigators and SPi i lI ti t d Sponsored d ProjectsSponsored Projects and ParticipantsAuthors and Publications
  27. 27. Adjacency MatrixA table of each node by each node A B C DA| x 1 1 0 AB| 1 x 1 0C| 1 1 x 0 B CD| 0 0 0 x D
  28. 28. Transforming Relational Database DataWhere your data has unique identifiers and features associated with them such as: them,
  29. 29. Transforming Relational Database DataJoin two instances of your table by the unique identifier:
  30. 30. Transforming Relational Database DataQuery for both instances of the feature, returning:
  31. 31. Transforming Relational Database DataNetwork analysis software will remove “self-loops” and duplicate edges: g
  32. 32. Transforming Relational Database DataAnd the resulting visualization might look like:
  33. 33. Unstructured TextNode: Word or phraseLink: Co-occurrence within a block of textSuppose we wanted to find co occurrences co-occurrences among words in unstructured text and words of interest included “network” and network “text.”You can construct a network based upon word co-occurrence in unstructured text text.
  34. 34. Unstructured TextYou can construct a network based upon word co-occurrence in unstructured text.Edge Listnetwork texttext network
  35. 35. Free, User-Friendly SoftwareNodeXL [] [ p p ]-Microsoft Research / University Collaborators Microsoft-Installs as an Excel 2007 Template-Free, easy, and powerful with top notch Free easy top-notch visualization
  36. 36. Free, User-Friendly SoftwareSimple Text/Network Mining p g-Homegrown Excel/Visual Basic Package-Tech Transfer [ - Search for # 4730]
  37. 37. LiveDemo
  38. 38. Specific ExamplesThings/Ideas and Other Things/IdeasConcepts and Other Concepts in Publications and Sponsored Project Proposal / Award Data
  39. 39. Concepts and Other Concepts in Publications and Sponsored Project Proposal / Award Data Concept C t Increasing Betweenness Centrality
  40. 40. Specific ExamplesPeople and Things/IdeasPeople and Sponsored ProjectsAuthors and Publication Concepts
  41. 41. People and Sponsored Projects Medical School PI Engineering PI Medical School Project Engineering Project
  42. 42. Specific ExamplesPeople and Other PeopleCo-Participation on Sponsored Projects, Co-Authorship
  43. 43. Co-Participation on Sponsored Projects, Co-Authorship Researcher / Author Active Project + Publication Increasing Ei I i Eigenvector C t lit t Centrality Active Project A ti P j t Publication
  44. 44. Strategies for CommunicationVisualization -Pay attention to node layout -Subtly encode as much d t as you can S btl d h data -Include a really simple keyYou understand the network data data, visualization, and metrics + your audience doesn t doesn’t = hand deliver
  45. 45. Q&A
  46. 46. Resources p p Transfer # 4730On Campus: School of Information, Center for Positive Organizational Scholarship Scholarship, Interdisciplinary Group for Research on Innovation