Text-mining for Medicinal Plants 15
Damien1972 writes "Researchers are exploring ancient texts for medicinal plant information using text-mining. From Shamans and Robots: Bridging the Past and Future of Ethnobotany and Bioprospecting: "A new procedure that is being explored by researchers to track and classify useful medicinal plant species may negate some of the issues surrounding the acquisition of knowledge ... This method involves a practice called "text-mining," in which old botanical works are scoured for references to medicinal plants.""
What about the ancient practice... (Score:4, Interesting)
As the saying goes; a month in the lab can save you a whole day in the library. Is this really new in a good way? Given the fact that many plants have alternate names, some have the same name, etc. it seems that familiarity with one's subject material is not a particularly useful thing to shortcut-out. Consider, for example, the parable of the 'mustard seed' in the gospels. A grown mustard plant is described as a great tree. Huh?! Even when people are familiar with the text, translations of old plant names are often difficult.
I'd just as soon read a machine translation of the bible or tartouffe than rely upon this technique. In fact, I'd sooner read a machine translation of the bible. Even if it mistranslated things, at least it'd be less likely to gloss the text. But that's not so much a problem with ethnobotany.
"Text mining" is a solution which found the wrong problem.