labels as words

[MarcoP]

As suggested by Davidsch, I have extended the label statistics to include the “Rosettes” page. I used Solfi's transcription for You are not allowed to view links. Register or Login to view. and Takahashi's for all the rest.

The percent results do not change with respect to those computed without the Rosettes:

539 total labels (single words with a minimum length of 4)

251 47% labels perfectly matching words
 23  4% labels matching a sequence of words
 38  7% labels also occurring as part of words
 43  8% additional labels found removing spaces
184 34% labels that cannot be found

Using these data, I have computed something along the lines of You are not allowed to view links. Register or Login to view., but maybe not as complex.
I made some double checking, but I cannot exclude I made some major error. If anyone thinks that these results are not consistent with what can be observed in the VMS, I would be happy to discuss the subject and make corrections!

I have split the data into sections according to the lists provided here:
You are not allowed to view links. Register or Login to view.
with the difference that I considered the Astrological and Cosmological to form a single section.

Then I have computed a graph describing the distribution with which labels appearing in all the sections appear as words in the other sections (so this is only about the 47% of the labels that perfectly match words).

In the graph, colors correspond to the sections containing the labels, while the bars correspond to the section containing the corresponding words. The bars represent percentage of the labels on each section: for each color, all bars of that color add up to 100%.
The herbal section contains only a handful of labels and the “stars with text” section none. So these two sections don't appear as label “sources”.



Comments:

• Labels re-occur as words rather uniformly in the different sections.
  
• The most prominent difference is that labels from the biological section tend to appear as words in the biological section more than elsewhere. I did not expect a single section to behave so differently from the others. But it must be said that the biological labels reoccurring in the biological section are only 34: the numbers are maybe too small to be statistically reliable. 
  
• Labels from a section seem to have a minimal but consistent preference to re-occur as words in that same section (the exception is the biological section, in which the preference is not minimal, see previous point).