BBC has used a words cloud to present the most frequent words in the presidential debate yesterday. While I think that words clouds can be fun and useful, I do not think it was the right tool to use in this case.
A word cloud can be useful to visualize which element is more important in a context but, to make this representation fair, the elements presented should not have competitors with the same meaning. Lets say it is used to present how many pictures you have, having a category animals and a category birds would make the comparison unfair ... now, imagine comparing from any word in English, what should happen with synonyms?.
Then how can this be useful? Maybe in two ways:
* Considering that candidate one used 7029 words and had less occurrences of the same (less words in the cloud) than candidate two who used 6546, we can assume that candidate one used a richer vocabulary.
* Now, considering that human brains and eyes are well trained in grouping words and trying to make sense from it, this visualization can result more useful as a mean of entertainment, like a "sentences soup". Can you find:
- "lot American people taxes free"
- " Colombia "(I didnt hear this one myself... )
- "http://www.cqpolitics..." (part of the debate ??? )
