Tag clouds are a neat way of summarising the common topics on blogs (the cloud on HolfordWatch is amusing, and this one for the Democratic race is neat). This week they’ve hit the science blogging world. The author of Clastic Detritus applied tag clouds to their papers. This is too neat an idea for me not to join in, so…

The tag cloud lists the most common words in a text, with the size being related to frequency. They can ignore the common English words like and and but. To produce these clouds, I used TagCrowd1

First up, a recent paper about Bayesian methods in quantitative genetics:

(hey, I didn’t say my papers were exciting)

Common English words are ignored, but I’m amused that et and al appear. Aside from them, I think it shows the main themes of the paper.

Next up, my anarchy paper:

I think you can work out fairly well what it was about. Similarly the buffeted paper:

is pretty clear.

And finally, for the moment, a paper in press in Biometrics:

Extra marks will be awarded to those who can work out what programmes were used to write the manuscript.

I might spend a few spare minutes getting these for all my papers, and putting them on the web. I’ve no idea if it’ll be useful, but it might be fun. It might also be a useful way of spotting clichés in your writing. Put several papers through the cloud generator, and see which words come up too often.

¹ Because NN can’t accept xhtml code, I clicked the “Printable/Full screen version” box, and then took a screen dump (PrtScr – top right of the keyboard) and imported it into a graphics package. I saved it as jpg, which is bad for this sort of graphic (try png instead), but I was uploading them to flickr, which would convert to jpg anyway.