I just finished my prototype last night
(I shall finished it earlier, but I am just too lazy~"~)

try this online

the data set is the abstract of the vis week papers during 1990 to 2008.


Sankey Diagram


偉大的Chris提出了一個超棒的想法:Sankey Diagram


然後他說我的東西結合Sankey Diagram大概就像Theme River結合Sankey Diagram





fetch related tags in different days of del.icio.us

For my study, I just write a small PHP program, which shows the relation of a specific tag in different days.

the php file could be download from here 

The return format is as follow
<data name="Queried Tag" >
   <date time="
YYYYMMDD" total="number of tag at that date">
      <tag name="
related tag" count="number of co-occurence"/>

Actually it directly parse the info on the del.icio.us page.
To avoid being banned by del.icio.us, the program would sleep 1 second for each 10 pages.

I would also think of adding the count of the related tag of a specific day. However, instead of really go through  all pages until find the day, I have no good idea of fetching those information. 

If any of you have a better solution, please help me, Thanks!!!!

Thanks for the HTML parser written by S.C.Chen.
For more detailed info, please visited the SourceForge
Also, there's a simple chinese tutorial over here


Difference between InfoVis and HCI

HCI focus on "New interaction skill"

Infovis uses tranditional interaction, but what may be more important is "New Layout"

User study is really important for HCI, and it also helps a lot on InfoVis.
(But I think they're not as important as that in HCI)