Joint Institute for Knowledge Discovery Exploring ENRON Email with NetLens Catherine Plaisant, Benjamin B. Bederson Hyunmo Kang, Bongshin Lee Human-Computer Interaction Laboratory University of Maryland
Our research focus Alternatives UI to Graph Visualization how to avoid this … Node-Link diagrams have many limitations. Not readable, may show clusters but not much else, do not scale well.
NetLens Iterative Exploration of Content-Actor Network Data � User Interface for exploratory search � Generalizable to a variety of data � Provide consistent interface � Easy to learn and use Kang et al. Proc. of Visual Analytics Science and Technology Conference (VAST 06) Kang and al. Poster/Demo at Joint Conference in Digital Libraries, 2006
NetLens Iterative Exploration of Content-Actor Network Data � Paired networks of Content and Actors, e.g. � Paired networks of Papers and Authors Papers refers to other papers � Authors have advisors � � Paired networks of Emails and People Email respond to or include emails � People have assistants who send email for them � � Paired networks of Products and Companies Products replace or integrate products � Companies are bought or merge �
Content-actor model Relationship Self-relationship Self-relationship Entity E1 Entity E2 Examples for scientific papers:
Toward SCALABILITY Total Enron email (non duplicate) 249,760 emails, 87,673 people
Email People (addresses) Overview by years Overview by Domain
Alternative overviews: emails by People by: day of the week, grouped by year connectance magnitude (Low medium high)
Multiple email search capabilities 1- Keyword Search Here a search on “ California ” 2- Similarity Search Find emails similar to one or more selected emails Result set loaded in “ My list ” (with Doug Oard ’ s team)
Social network analysis: - Number of neighbors - Connectance - Centrality - Average Path Length - Here selected people with high connectance With Jen Golbeck
Social network analysis: - Number of neighbors - Connectance - Centrality - Average Path Length - Here selected people with high connectance With Jen Golbeck Explanations of the meaning of the attributes
People bios � Using signatures and directory info with Jen Golbeck �
Integrated Phone calls Replay Separate conversations Direct access to mentions of : subject, names, keywords (with Carol Espy ’ s team)
Thread Summaries -List of emails in same thread -Access to thread -Access to thread summary With Bonnie Dorr and Doug Oard ’ s teams
TreePlus to browse subset of network connections
TreePlus - Visualizing Graphs as Trees � Plant a seed and watch it grow � Faster, more accurate, preferred over traditional graphs for tasks that involve reading and exploration of connections � To show hidden graph structure � Highlight and preview of adjacent nodes � Animated change of tree structure � Visual hints about graph structure B. Lee, C.S. Parr, C. Plaisant, B.B. Bederson, V.D. Veksler, W.D. Gray, C. Kotfila (2006) TreePlus: Interactive Exploration of Networks with Enhanced Tree Layouts To appear in TVCG Special Issues on Visual Analytics B. Lee, C.S. Parr, C. Plaisant, B.B. Bederson (2005) Visualizing Graphs as Trees: Plant a seed and watch it grow Proceedings of GD 2005 (poster) , LNCS , pp. 516-518
Generalization to other datasets e.g. NetLens for Scientific Publications (Papers and Authors)
User evaluation � Heuristic review at NIST � 5 PEOPLE – self trained with video) � Usability Study � 9 people, training, debriefing � Other improvements Improved feedback � • +++ Improvement of flow management Addition of My List � Adaptive explanations of views � Video training � Documentation of source / processing of variables �
Implementation � C# (using piccolo toolkit) � MS Access Database � NetLens component code available on request
Conclusions - Future Directions � Conclusions � Simple content actor model helpful � Powerful yet simple � Training about flow behavior � Continue integration with other IJKD data � E.g. Entity resolution � Evaluation (case studies of analysis) � Needs for Proto � Tool � Facilitate code customization for different applications � Flexible entities switching (to handle any choice of pairs) � Usability
Thank You � plaisant@cs.umd.edu (301)405-2768 � bederson@cs.umd.edu (301) 405-2764 � NetLens: www.cs.umd.edu/hcil/netlens � TreePlus: www.cs.umd.edu/hcil/treeplus � Papers and Video demonstrations available from website. � Source code available on request.
OTHER relevant HCIL projects
Temporal Data (Categorical): PatternFinder for Patient History Search Fails, Karlson, Shahamat & Shneiderman, VAST 2006
Systematic & Flexible Network Exploration with SocialAction Abstraction reveals Clustering shows relationships grouping Perer & Shneiderman, InfoVis 2006
Network Visualization with Semantic Substrates • Meaningful layout of nodes • User controlled visibility of links Shneiderman & Aris, InfoVis 2006
Recommend
More recommend