cs6630 | October 23 2014 TEXT & SETS Miriah Meyer University of Utah administrivia . . . 2 - parallel coordinates due next Thursday 3 last time . . . 4 5 GRAPHS & TREES - - graphs - model relations amount data - nodes and edges trees - graphs with hierarchical structure - nodes as parents and children 6 7 http://bl.ocks.org/mbostock/4339184 8 http://bl.ocks.org/mbostock/raw/4348373/ 9 http://hci.stanford.edu/jheer/files/zoo/ VISUALIZING GRAPHS - node link layouts Reingold-Tilford (trees only) - Sugiyama (directed acyclic graphs) - Force directed - Attribute-based - - adjacency matrices - aggregate views Motif Glyphs - PivotGraph - Les Misérables character co-occurrence 11 http://bl.ocks.org/mbostock/4062045 Les Misérables character co-occurrence 12 http://bost.ocks.org/mike/miserables/ MOTIF GLYPHS 13 Dunne 2013 PIVOT GRAPHS - new graph, derived from categorical node attributes - 1D or 2D layouts possible - - size of nodes and edges related to number of aggregated original nodes and edges scalability through abstraction, not layout algorithm 14 today . . . 15 16 Text 16 TEXT 17 Text ? what does it mean to be an “item”? 18 text data type ! - no numbers (implicitly) ! - characters: ASCII ! - strings 19 text data type ! - no numbers (implicitly) ! - characters: ASCII ! - strings 19 text data semantics e v lo ! - words ! - visu aliz atio n sentences - paragraphs - chapters ! - lines 20 I . text data semantics e v lo ! - words ! - sentences - paragraphs - chapters visu aliz atio n I love visualization. ! - I lines 20 . text data semantics e v lo ! - words ! - sentences - paragraphs - chapters visu aliz atio n I love visualization. ! - I lines 20 . text data semantics ! - documents - books - papers - webpages - emails - twitter post ! - corpus: collection of documents 21 text data semantics ! - documents - books - papers - webpages - emails - twitter post ! - corpus: collection of documents 21 22 22 single document Tag Clouds / Word Clouds http://www.tagcrowd.com http://www.wordle.com Text Arc Wattenberg, Viegas 2008 DocuBurst Collins, Carpendale, Penn 2008 Arc Diagrams Analysis of the Characters from Les Misérables: http://mbostock.github.io/protovis/ex/arc.html Rule-Based: Poetry Abdul-Rahman et al. 2011 collection of documents Document Cards (small multiples) Showing Temporal Relationships: ThemeRiver (Stream Graph) Havre, Hetzler, Nowell 2000 Jigsaw: Many Linked Views Stasko et al. 2008 Jigsaw: Many Linked Views Stasko et al. 2008 SETS 35 Text ? 36 thought experiment… - item: Lego - attributes 37 thought experiment… - - item: Lego attributes - color - height - width - length - shape 37 dataset: option 1 ! 38 dataset: option 2 ! 39 dataset: more realistic ! 40 dataset ! - where do we start? we need to organize! but, how? 41 dataset ! - sort by color 42 dataset ! - sort by size, shape 43 dataset ! - task: organization ! ! - drawbacks? 44 dataset ! - organization leads us to a set problem ! - so what are sets? 45 set theory ! - set - a collection of objects - some set: A ! - A z object - some object: z -z ∈ A 46 set theory ! - set - a collection of objects - some set: A 47 A set theory ! - multiple sets: A & B A B 48 set theory ! - union: A ∪ B A B 49 set theory ! - intersection: A ∩ B A B 50 set theory ! - set difference: A \ B A B 51 set theory ! - symmetric difference: A ⊖ B A B 52 visualizing sets 53 venn diagrams ! - show all possible relationships 54 venn diagrams ! - casual infovis 55 venn diagrams ! - get messy fast 11 7 56 venn diagrams ! - non-sensical 57 euler diagrams ! - show only existing relationships 58 euler diagrams ! - show only existing relationships 58 euler diagrams ! - misunderstood 59 euler diagrams ! 60 venn & euler diagrams ! - adjust for area ! ! - starts getting tricky! 61 venn & euler diagrams ! - adjust for area ! ! - starts getting tricky! 62 compact euler diagrams ! 63 parallel sets ! 64 set o’gram ! 65 visualizing sets with constraints 66 bubble sets ! - connect points 67 line sets - restaurants social communities 68 kelp diagrams ! ! ! ! - cities on a map ! - metabolic network 69 kelp fusion ! - cities on map ! ! - lines & areas 70 sets ! - applies to many datasets - many combinations may be interesting - limited numbers of sets Text ? 71 L15: Maps REQUIRED READING 72