I’ve posted a brief tutorial on using JWNL, a Java library for accessing WordNet.
“WordNet is a large lexical database of English, developed under the direction of George A. Miller. Nouns, verbs, adjectives and adverbs are grouped into sets of cognitive synonyms (synsets), each expressing a distinct concept. Synsets are interlinked by means of conceptual-semantic and lexical relations.”
I wonder how useful it would be to create a Processing library with some basic WordNet functionality (or that simply exists as an interface to JWNL for all things WordNet.) There are many lovely examples of text-based Processing work out there. Two that jump to mind are State of the Union by Brad Borevitz as well as anything ever made by Ariel Malka.
The power of WordNet is not that it provides access to word meanings (which it does), but that it provides information about the relationships between words (and when I say “word,” I really mean set of synonyms, or synset). It’s the myspace for semantic concepts. You’re probably familiar with synonyms and antonyms. In addition, WordNet provides links for less well-known semantic relationships, such as hypernymy, hyponymy, meronymy, troponymy, and entailment. Do I see some sort of tree / network visualization in Processing’s future?
Hi Daniel,
Way back in grad school I did something along those lines; you can read a paper about it here: http://www.hml.queensu.ca/papers/po265-senior.pdf
I had to calculate the semantic relationships myself, using co-existence in Google search results as a metric for closeness. The viz of the system shown in the paper is quite boring — I’d love to have another crack at it using WordNet and Processing.
I originally started using Processing for the viz, but it wasn’t stable enough at the time so I had to scale back my plans.
Cheers,
-robin