pcplantdb

[Bear Kaufmann <bear@ursine-design.com>]

> If feedback on any of this from nonprogrammer PIW/PCPlantDB afficianados
> (in Flamenco, these are the folks who keep up with the world of Flamenco and 
> all the notable exponents of that art,
> learn the forms of song and dance, go to juergas, drink brandy, stay up late 
> and generally love that music)
> would be useful here then could you post a little explanation of what you're 
> trying to do and define some of the
> terms involved? I.e. what is text analysis about in the context of a pc/relationships database? Does it have to do with 
> analysis of the text output of a DB search or to do with parsing search strings entered in the DB query form?
>  
Sorry, some of these discussions occurred off list.
I went through the process of dividing all of the comments in the PIW 
dataset into individual words. Each word is also identified with the 
plant it was used on. From this, we can do a number of interesting things.

Here are the most common words in the dataset (top 3,000 out of 16,000):
http://permaculture.info/test/most_common_words.html

We can find what words co-occur often (appear within the same plant 
description together):
http://permaculture.info/test/related_words.php

If you search for 'woven' in the related words, you'll see a reasonable 
example.
It includes: layers thatch chairs insulating nappies sheets blankets
'Ice' includes: cream sweets creams canada spearmint
'Fronds' includes: ferns spores carcinogens humid fern plantlets bracken
'Stronger' includes: oregano savory onions crispum hortensis pepo 
pumpkins affecting beet chamomile curly digitalis dogs gutta-percha 
moles onion marjoram marrows.

You can also apply something similar to what google does when you use 
'find similar' on a page:
http://permaculture.info/test/related_Abutilon_vitifolium.html
http://permaculture.info/test/related_Abelmoschus_esculentus.html
http://permaculture.info/test/related_Acer_crataegifolium.html

Note that none of the above results had anything to do with the name of 
the plant, only the shared words. The shared words appear below the 
plant name.

All of this is non-semantic, meaning that "not edible" and "edible", 
both match edible. "not eaten by squirrels" gets 3 hits for "eaten by 
squirrels". But you still get interesting results.

I'm also right now working on something using google: getting the number 
of hits each plant has on the web. Right now Aloe vera is the clear 
winner, but I've only done genuses from A-C, since I don't want to 
overwhelm google's servers. Second highest currently is a model species 
for plant biology, then food crops, herbs, and ornamentals. I'm curious 
what the differences between google general and google scholar would 
show. Those plants with low scores generally place PFAF and PIW as the 
top hits.

-
Bear