I'm working in a research project:
a Deaf'n'Dumb translator...
It built in java calling Matlab and Cyc expression!
Taking some fotograms, I segmented the Hand,then I passed the to Som-NeuralNetworks that build a table of probability.Example:
First word | Second Word | Third Word
90% Me | 100% TOGO | 90% NO.
5% Day | | 10% Cinema.
So Passing all the words, I've to build a Linguistic analysis to determinate in my microtheory in Cyc The sense-sentence... I know It's a little bit hard to understand... but It's my work... How do you think it's the better way to make this Mtheory?
The Deaf'n'Dumb language is more easy than English... there aren't articles... propositions... and the words I use are almost 100.
Could you help us by listing the 100 words that are in the Deaf'n'Dumb language?
It might make it easier to see how to tie that to an OpenCyc kb.
I realized that Deaf'N'Dumb language isn't Natural-Language... but something like this... so the problemsolving may be easier than I think....
with Java or AIML I build something like a Chatbot (e.g. ALICE) and with my Microtheory form Natural LanguageMt (with words already are there... not lo listing from scartch ). I verify if the sentence has sense looking if the words in OpenCyc kb are Verbs,Noun... and in Java if there are in the right position (e.g. Noun before Verb).
Maybe now It works...
Tell me your impressions!
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.