Scholarly journal style 2.0 (Riff on the wooden table theme)



  • This is really of marginal relevance to WTF, but the fact that the "wooden table" joke never seems to go stale here gave me courage to post.

     

    The authors of Spangler W.S., et.al., "Machines in the conversation:  Detecting themes and trends in informal communication streams", IBM Systems Journal, vol 45, no 4, pp 785--799, had some reasonably small datasets to visualize in their paper.  For example, one graph (figure 3) had word frequency values for four different words in two different contexts for a total of 8 values.  Not content with a simple bar graph, they:

    1. created 3-D bars
    2. rotated them out of the plane of the axes
    3. surrounded the resulting graph with a fake "window" complete with scroll bar (a big help on the printed page !!)
    4. rotated the entire figure and pasted it into a simple drawing of a flat-screen monitor.

     

    graph with computer monitor 

     

    (entire article available http://www.research.ibm.com/journal/sj/454/spangler.html online) 



  • Well that's sure a whole lot easier and better looking than taking a screenshot of excel...



  • Argh, the angle of the bars and the angle of the monitor combine to give me a headache! I think I just had a coredump in my brain, I wonder if this is remotely exploitable.....

    But seriously, my question is..... where's the wooden table? :P
     



  • This is often found in magazines and newspapers. But in a research paper?

     
    And it doesn't look good, either. Looks like it was done using the drawing features from Word.



  • That's fantastic. They even went to the trouble of putting a (totally meaningless) scrollbar in the window, but didn't think to put a title bar on it. For some reason, a group box with one child control has become the window's title bar and border. The slanty angle of course is nuts.

    Awesome.



  • I like how they make use of not one, not two but three different 3D systems, each with its very own set of angles:

    First there is the "flat screen", complete with skewed window content

    Second there are the 3D features of the fake scroll bar and the border

    Third there is the graph itself...

     

    Talk about consistency...
     


Log in to reply