Slashdot Mirror

← Back to Stories (view on slashdot.org)

Software for Social Networking Diagrams?

Posted by Cliff on from the we-could-use-Twin-Peaks-as-a-test-case dept.

jonerik writes "A friend of mine who is attempting to create three-dimensional computer models of social networks is looking for software that can do the job; in her words, 'the computer version of a tinker toy set.' She'd prefer software for the Mac, however she has access to PCs so MS-based software would be okay, too. Anyone have any thoughts on the subject?" Think about molecular modelling, where each "atom" is either an agency or an individual and the "bond" is the connection between the two.

"More specifically, here's what she'd like it to be:

  • easy to learn for someone without graphics background- 'I can NOT make a career out of learning it.'
  • something like a molecular model of social networks where each node represents an agency and can have many different types of connections. It should at least be able to handle something on the order of 250 interconnected nodes.
  • able to edit the model on the fly - adding information as I learn about connections.
  • able to rotate the model, enlarge sections to view detail, and maybe somehow attach descriptions to the nodes and connections (so one can identify a connection with a particular agency role or individual)."

11 of 24 comments (clear)

  1. The molecular model is too simple. by Futurepower(R) · · Score: 4, Insightful


    To better explain my comment above: The molecular model is too simple. There are many more different kinds and qualities of connections between people than kinds of molecular bonds. So, software to display connections between people would need a far more complex display.

    My guess is that no such software exists. This is, apparently, a huge new programming problem. It is also one that would be very valuable to solve.

    I did a sociogram of myself once, and discovered that I knew about 200 people (some of them secretaries with whom I worked, and so on). Even just drawing simple lines between people was interesting.

    The interesting challenge would be, not to show relationships in a static way, such as "romance", but to show dynamic interaction, such as "B influences A in ideas about psychology".

    1. Re:The molecular model is too simple. by The+Mayor · · Score: 2

      It seems I remember reading in a Wired issue a few months back about software the CIA was using for this purpose. Apparently, they are using the software to develop an idea of terrorist networks, the way they operate, and their command structure.

      I believe the software packages the CIA uses are either developed internally or developed specifically for the CIA (read: not for public consumption). Sorry I can't give you any more details on this. It's been a while, and my memory isn't so good.

      --
      --Be human.
    2. Re:The molecular model is too simple. by XoXus · · Score: 2
      Yeah, but I'm sure that the CIA would be more than happy to open-source this kind of software...


      Hehehe... it'd be funny if they actually did, though.

  2. Links by Anonymous Coward · · Score: 2, Informative
    Node Graphs? The answer is Graphviz. It is not oriented to 3D but handles the difficult layout problem well. Definitions are in a simple text format that I've found I can easily create from many datasets.

    You mentioned molecular figures, so I assume you already looked at software for creating molecular stick figures. Not much, is there?

    Gnuplot has amazing capabilities. It's not oriented to 3D realtime animation, but you can get 3D plots...and make animated PNGs to rotate around a plot. Ignore the function abilities, look at the point and line plotting.

    Did you do a search for scientific visualization? If you had, one of the things you'd have found is Geomview from the U of MN Geometry Center. Don't know how easy you'd consider it.

  3. Maybe Intermapper by Evro · · Score: 2, Informative

    I'm not totally clear on what she wants here, but check out Intermapper. It's sort of cool even if it's not what you're looking for.

    --
    rooooar
  4. Pajek by romanm · · Score: 2, Interesting

    Your friend can try Pajek, a package for large network analysis for Win32, developed at University of Ljubljana, Slovenia. Or you may want to search the Net for other Graph Drawing Competition contestants.

  5. Data Explorer by scotpurl · · Score: 2

    http://www.opendx.org/, a now open source product from IBM. Does wonderful 3-D stuff. Source code available, and binaries available for AIX, FreeBSD (might compile on OS-X), HPUX, Irix, Linux, Linux PPC, Solaris, and Microsoft OS.

    Be sure to check out the highlights: http://www.opendx.org/highlights.php

    Using this software won't be point, click, drag, like using Visio, but I'm sure you could easily compose the modules that would parse data, and create what you need. Look at the samples and highlights on the web page before you decide.

  6. Think RDF by Bazzargh · · Score: 4, Informative

    RDF - the XML format for metadata - essentially describes relationships between resources and is a superset of the problem area you describe. I had a cursory look to see if people had taken the obvious step of using diagram tools to edit rdf and yes they have:

    Using DIA and GraMToR and GraphViz and IsaViz and lots lots more

    You might also be interested in They Rule which draws some interesting social networks!

    While this is all standard 2d editing (I think the 3d editing thing is nonsense btw, unless you actually have a 3d coordinate system then you are actually describing a 2d network diagram) there are other visualisation options which may be interesting. Mindmapping tools allow you to navigate the network of concepts may be interesting as are star trees. Both provide a focussed view on a small part of a much larger network with some context information to help you choose how to navigate. They are more useful for display than authoring.

    Hope something in here is useful

    -Baz

  7. the sexchart by Phork · · Score: 3, Interesting

    the sex chart is a 2d rendering of 1500 iinterconnected nodes. I belive they have a utility that gets fed a textfile describing the node relationships, and from that reners the text. This might be usefull to you.

    --
    -- free as in swatantryam - not soujanyam.
  8. network mapping software by orgnet · · Score: 3, Informative

    3D molecules? Try Mage: http://kinemage.biochem.duke.edu/

    Warning: these 3D diagrams lose their appeal when the network > 40 nodes. 2D still works the best for large networks.

    A few real network maps...

    • 9-11 Hijackers & associates [63 nodes]: http://www.orgnet.com/hijackers.html
    • Internet industry partnerships [222 nodes]: http://www.orgnet.com/netindustry.html
    • Interactive Java applet of Internet industry partnerships [try right-click options]: http://www.orgnet.com/inetindustry.html
    • Advice network in I/T department[64 nodes]: http://www.orgnet.com/experts.html

    And finally, a brief intro to Social Network Analysis: http://www.orgnet.com/sna.html

    Enjoy!

  9. Visualizing Social Networks by nobodaddy · · Score: 2, Informative

    Visualizing Social Networks
    http://zeeb.library.cmu.edu:7850/JoSS/ar ticle.html

    State of the Art of Computing in Social Science
    http://www.cni.org/Hforums/gils/1999/0018 .html

    Social Networks @ arizona.edu
    http://w3fp.arizona.edu/soc/areaDetai ls.asp?tblTex t_category=networks

    A Guide for the Visually Perplexed: Visually Representing Social Networks
    http://www.stanford.edu/group/esrg/silic onvalley/d ocuments/networkmemo.doc

    go wild.