TileBars

By: Jeff Clark    Date: Thu, 30 Nov 2006

TileBars are small rectangular graphics that represent documents and illustrate the frequency and distribution of sets of related terms across the documents. The concept was developed in the 90's by Marti Hearst, currently at UC Berkeley.

Here is an example:



 
The differing lengths of the bars illustrate the relative lengths of the documents and each row of small squares is shaded to show the frequency of the corresponding termset in that region of the document. In Marti's words:

The goal is to simultaneously and compactly show:
  • the relative length of each retrieved document,
  • the frequency of the topic words in the document, and
  • the distribution of the topic words with respect to the document and to each other.

I like this idea but it hasn't seemed to have made it's way into popular use.

 


Digg Topic Popularity over Time
Blog
EagerEyes