bcholmes: (i am not a number)
[personal profile] bcholmes

Imagine this scenario: you have a large body of newspaper articles. You can extract from all of the articles a list of key words. And you can gather metrics about which key words are most common, etc.

A tag cloud is a traditional way of representing the relative importance of a list of key words. What a tag cloud lacks is any representation of the relationship between key words. If you can say that "AIDS" and "HIV" are very frequently used together, but that "Pope" is only occasionally used in the same article, is there a good visualization that shows both relative importance and relationship?

(no subject)

Date: 2010-11-24 06:51 pm (UTC)
elainegrey: Inspired by Grypping/gripping beast styles from Nordic cultures (Default)
From: [personal profile] elainegrey
IBM's Many Eyes project has some interesting textual data visualizations. This example shows a visualization that might be better for your purpose

http://www-958.ibm.com/software/data/cognos/manyeyes/visualizations/i-wish-part-2-3-key-phrases

Many Eyes is great fun to use; you just need data you don't mind the world seeing.

Profile

bcholmes: (Default)
BC Holmes

February 2025

S M T W T F S
      1
2345678
9101112131415
16171819202122
2324252627 28 

Most Popular Tags

Page Summary

Style Credit

Expand Cut Tags

No cut tags
Powered by Dreamwidth Studios