I have generated Topic Flowers for the current top 2 articles in Google News for the categories Business, Sci/Tech, Sports, and Health. Here they are with links to the articles and their topic as defined by Google:
1 | Business - | I'll be back, says fallen Wall Street dotcom star | |
2 | Business - | Smoke in cabin forces plane evacuation | |
3 | Sci/Tech - | Astronomers offer proof of unseen 'darkmatter' | |
4 | Sci/Tech - | Pluto Seems Poised to Lose Its Planet Status | |
5 | Sports - | Yankees finish historic sweep of Red Sox | |
6 | Sports - | U.S. opts for new formula | |
7 | Health - | Chemical Reaction May Have Caused Eye Infections | |
8 | Health - | Accutane acne drug raises cholesterol, fat levels |
Using the topic legend we can interpret the colours. The flowers from the same categories as defined by Google do seem to be similar to each other. The business stories, numbers 1 and 2, are predominantly purple as expected. The Sci/Tech stories, numbers 3 and 4, are mostly blue. The two sports stories, 5 and 6, are all orange, and the Health stories, 7 and 8, are blue for science content with a little bit of purple for business mixed in. It is a little hard to see because I made the images smaller but most of them have very few 'hairs'. This suggests they don't contain many personal pronouns which makes sense for these articles that are from newspapers. They all have many layers of petals which shows that they are all reasonably long.