network visualization introduction
play

Network Visualization Introduction Presented by Shahed - PDF document

Network Visualization Introduction Presented by Shahed Introduction Introduction Basic building blocks Node Links (relationship between nodes) Spatial information Network data


  1. Network Visualization Introduction Presented by Shahed Introduction Introduction • Basic building blocks – Node – Links (relationship between nodes) – Spatial information – Network data http://zeeb.library.cmu.edu:7850/JoSS/article.html 3 4 Paper List Paper List • Visualizing Network Data • Visualizing Network Data – Richard A. Becker, Stephen G. Eick, Allan R. – Richard A. Becker, Stephen G. Eick, Allan R. Wilks. Wilks. • 3D Geographic Network Displays • 3D Geographic Network Displays – Kenneth C. Cox, Stephen G. Eick, Taosong He. – Kenneth C. Cox, Stephen G. Eick, Taosong He. • CyberNet: A framework for managing • CyberNet: A framework for managing networks using 3D metaphoric worlds networks using 3D metaphoric worlds – P. Abel and P. Gros and D. Loisel and C. Russo – P. Abel and P. Gros and D. Loisel and C. Russo Dos Santos Dos Santos 5 6 1

  2. Goal Traditional Approach • To reduce cluttering of data (traditional) • Visualize the data associated with a network – Aggregation : for large numbers of links or nodes – Understand data, not network themselves • Coping with large data volumes – Averaging : for large numbers of time periods – Hundreds of nodes – Thousands of links – Thresholding : for detecting changes – Data from time periods • Overcome the map clutter problem 7 8 Solution Dataset • SeeNet • Telecommunication traffic – Static Displays • 110 switches in the AT&T network • Link Map • 12,000 links • Node Map • Matrix • Oct. 17, 1989, (San Francisco – Interactive Controls earthquake) • Parameter focusing • FOCUS: • Data filtering – Animation – Traffic flow between switches (nodes) • Smooth zoom 9 10 Static Displays (1/3) • LINK MAP Static Displays – Draw lines connecting nodes – Show values using colors or thickness of line 12 2

  3. Static Displays (LinkMap) Static Displays (LinkMap) Focus on one Node (Oakland ) Include all nodes (10% of links shown) 13 14 Disadvantage Static Displays (2/3) • Disadvantage of Link Map • NODE MAP – Too many links cause map cluttering – Aggregation of information at each node – Use Node Maps !!! – Use Glyphs • Vary Size, shape, color for statistics 15 16 Static Displays (NodeMap) Disadvantage 1) Tall & Thin: Outbound • Disadvantage of Node Maps overload (green) – Detailed Information about particular links lost 2) Short & Fat: Inbound Overload – Solution: (red) • Do away with geography • Try Matrix display 3) Square: Equal load (white) 17 18 3

  4. Matrix Display 19 20 http://funwavs.com/movie/pictures/the-matrix/ Static Displays (3/3) Disadvantage • MATRIX • Disadvantage of Matrix Display – Concentrates on links of a network (like Linkmap) – Information about geography lost – Color of square designates traffic – Does not have problems of geographic • Tries to fix problem with nodes ordered from west coast to east coast along axis displays: • Visual prominence of long lines • Long lines (transcontinental) over plots others 21 22 Parameter Focusing Parameter • Parameters determine network display Focusing • Parameter values (range) control what is displayed – Example: – Glyph size in node maps – Coloring of nodes & links • Dynamic parameter adjustments helpful 24 4

  5. Parameter Focusing Example Parameter classes (Shortened Links) • Statistics • Levels • Geography / topology • Time • Aggregation • Size • Color 25 26 Issues with parameter Direct Manipulation focusing • Automatic animation • Space of parameters large • Manual animation • Combination of parameters to chose • Sound • Displays sensitive to particular • Conditioning (‘and’ operation on parameter values parameters) • Identification (display tool tip of node) • SOLUTION • Zoom – Allow Direct manipulation of parameters • Birds-eye view 27 28 Other applications (non Example (zoom in Link Map) geographic) • Left: All line segments intersecting the display • Middle: any line segments with at least one endpoint in the display • Right: only lines that both begin and end inside the display CICNET EMAIL Communication 29 30 5

  6. Critique Paper List • The Good • The Evil • Visualizing Network Data – Clear graphs with – Self evaluation (no – Richard A. Becker, Stephen G. Eick, Allan R. interpretation user studies) Wilks. – Presented motivation – Redundant • 3D Geographic Network Displays and challenge information – Kenneth C. Cox, Stephen G. Eick, Taosong He. papers (parameters and – Tested on different • CyberNet: A framework for managing direct manipulation) data sets networks using 3D metaphoric worlds – Provides – P. Abel and P. Gros and D. Loisel and C. Russo implementation Dos Santos details (C++ & Vz) 31 32 Introduction Global Network 1/2 • Presents “ SeeNet 3D” – 5 network views • 2 views are geography related • 3 views concentrate on portion of a large network • SeeNet3D follow-up of – SeeNet – NicheWorks Global packet count in 2 hour period Tall red glyphs have more traffic 33 34 Global Network (2/2) Disadvantage - Arc Maps • Only Front side of map viewable • Draw arcs on flat 2D map in 3D space – Occludes arc ends • Solution – 2D map can be oriented as desired – Make globe partially translucent (does not work with too many arcs) – Eliminates line crossing to a certain extent (vary arc height) – Allow user to route arcs (through globe if needed) – Filtering 35 36 6

  7. Arc Map without Arc Map with parameterization parameterization of height of arc height Add translucency of arc &, coloring and size glyphs of countries 37 38 Drill Down network views Spokes on a wheel • Three types of views: -Works for 50 to 100 nodes -Does not make efficient – Spokes on a wheel use of screen space : All spokes of equal length – Helix -Better approach (Helix) – Pin Cushion Traffic to/from US to other countries 39 40 Helix Pin Cushion -Motivated by helix -When Viewed from above, display Helix view becomes spoke view -Position uniformly around sphere (anchor -Use rotation of helix to node) bring occluded nodes into view - Number of circles and number of nodes per -Preferred approach by circle chosen such that authors over others (more angle between circles ordered) and between nodes in a circle same 41 42 7

  8. Critique Paper List • The Good • The Evil • Visualizing Network Data – Shows more in less – No user studies (as – Richard A. Becker, Stephen G. Eick, Allan R. space (5 vs 3) usual) Wilks. – Provides – Some displays have • 3D Geographic Network Displays implementation limited information – Kenneth C. Cox, Stephen G. Eick, Taosong He. details – Does not give • CyberNet: A framework for managing scalability constraints networks using 3D metaphoric worlds for most – P. Abel and P. Gros and D. Loisel and C. Russo Dos Santos 43 44 Geographic administration Introduction building metaphor -For physical link problem detection • Network administration in 3D • Provides 5 metaphors -Building ( a container for network • Dynamically builds & updates 3D world devices) • Captures information -Object location is relative to position in – Topology, Connectivity, Routing, Mailing, actual world NFS -User allowed to • Each 3D tool solves specific problems – chose destination chose metaphor that best suites a task (automated paths) -Filtering 45 46 Distributed system admin Topology administration cone-tree metaphor city metaphor - Maps Client/server Red: switches (Mail, DBMS, NFS) Blue: Hubs Leaves: Computers -Separate Client & server view Size of cone depends on bandwidth flow in -Metaphors: hub Town : sub network District : Computer Building : Disk resource On server: -Each client a floor -Each window a 47 48 File Handle 8

  9. Computer admin tool Network traffic characterization solar system metaphor landscape metaphor Metaphors: Stars Planets satellites To: Computers Users Processes 49 50 More on CyberNet Critique • Users can toggle between various 3D • The Good • The Evil structures – design architecture – Some concepts explained unclear (city • Technical Stages metaphor) – Implementation – Collecting Layer (subscribe/notify, agents) language for each – No user studies stage (VRML, corba, • Mentions users found – Structuring Layer (build service model tree) Java, perl) metaphors helpful – Visualization Layer (generate 3D form) – No scalability – Screen Shots helpful discussion – Dead Site !! 51 52 ? QUESTIONS ? 9

Recommend


More recommend