Download (English) millenium-eng.pdf (215kB).
Download (Spanish) millenium-esp.pdf (235kB).

Based in the Stieg Larsson Bestseller. All the words from the Millenium trilogy.

In this case the problem was the huge step between the first words. Take a look at the word list:

blomkvist 2521, salander 2376, police 879, vanger 878, berger 703, zalachenko 589, murder 584, lisbeth 526, investigation 475, apartment 468, harriet 453, bjurman 448, bublanski 405, mikael 396, armansky 379, millennium 356, security 345, coffee 322, martin 319, computer 309, wennerström 307, ekström 302, svensson 300, henrik 288, niedermann 287, spent 286, taken 284, giannini 270, photograph 268, desk 267.

In only 30 words we have tenth of the initial size. The words becomes like ants quickly.

So I decided to insert a kind of ‘brake’ in the algorithm that makes the steps between words shorter. Yes, we lose accurate proportionality but still retain importance degree between words.

Images free for personal use only.

The spanish version uses the spanish books. No google translation.

Download house-eng.pdf (234kB).

Download house-esp.pdf (264kB).

Hands on it, I collected all the subtitles of the first five seasons from the website http://www.tvsubtitles.net/ something like 115 episodes.

Because the subtitles are ‘common language’ there is not so much information as in a song or a speech. I needed to clean up the word list in order to have a significant result.

I also realized that there is a lot of strange words with only one hit, probably due to the nature of the subtitles. I decided to include only the words with two or more hits.

Here is the final image!… Cuddy would be proud.

As usual, all files for personal use only.